INFORMATION SYSTEMS IN MANAGEMENT

Transcription

1 INFORMATION SYSTEMS IN MANAGEMENT Systemy informatycne w arądaniu Vol No. 2 Quarterly

2 Information Systems in Management Primary version of the journal is an electronic version. Editor Department of Informatics, Warsaw University of Life Sciences SGGW Editorial Committee Dr hab. inż. Arkadius Orłowski Editor-in-Chief Dr inż. Piotr Jałowiecki Scientific Secretary / Executive Editor Prof. nadw. Kris Gaj Linguistic Editor Dr hab. Wiesław Scesny Statistical Editor Dr Piotr Łukasiewic Technical Editor Editorial Council Dr hab. inż. Arkadius Orłowski Skoła Główna Gospodarstwa Wiejskiego w Warsawie Chairman Prof. dr hab. inż. Rysard Budiński Uniwersytet Sceciński Prof. dr hab. Witold Chmielar Uniwersytet Warsawski Dr hab. inż. Lesek Chmielewski Skoła Główna Gospodarstwa Wiejskiego w Warsawie Dr hab. Maciej Janowic Skoła Główna Gospodarstwa Wiejskiego w Warsawie Dr inż. Waldemar Karwowski Skoła Główna Gospodarstwa Wiejskiego w Warsawie Prof. Yuiry Kondratenko Black Sea State University, Ukraina Prof. dr hab. Marian Niedźwiediński Uniwersytet Łódki Dr inż. Jery Pejaś Zachodniopomorski Uniwersytet Technologicny w Scecinie Dr hab. Marian Rusek Skoła Główna Gospodarstwa Wiejskiego w Warsawie Dr hab. inż. Ludwik Wicki Skoła Główna Gospodarstwa Wiejskiego w Warsawie Prof. dr hab. inż. Antoni Wiliński Zachodniopomorski Uniwersytet Technologicny w Scecinie Address of the Editor Faculty of Applied Informatics and Mathematics, WULS SGGW ul. Nowoursynowska 66, Warsawa, Poland [email protected], ISSN: Wydawnictwo SGGW ul. Nowoursynowska 66, Warsawa, Poland [email protected], Print: Agencja Reklamowo-Wydawnica A. Gregorcyk,

3 INFORMATION SYSTEMS IN MANAGEMENT Vol No. 2 Table of contents Rysard Antkiewic, Michał Dyk, Rafał Kaspryk, Andrej Najgebauer, Darius Pierchała, Zbigniew Tarapata CRIMINAL PROCEDURE MANAGEMENT BASED ON BPM SIMULATION Adam Deptuła GAME-TREE STRUCTURES WITH THE COMPLEX COMPLEXITY LEVEL AS A TOOL IN KNOWLEDGE ENGINEERING Viera Gafrikova, Zdisław Odrygóźdź, Wiesław Scesny APPLICATION SUPPORTING COMMERCIAL BANK S CAPITAL MANAGEMENT... 3 Grażyna Golik-Górecka DASHBOARD CONCEPT AND APPLICATION IN THE MEASUREMENT OF MARKETING EFFECTIVENESS Artur Hłobaż, Kuba Owsiński PROPOSAL OF A NEW METHOD FOR E-BANKING AUTHENTICATION BASED ON FINGERPRINT READER Mirosława Lasek, Symon Jessa CHATBOTS FOR CUSTOMER SERVICE ON HOTELS WEBSITES Tomas Rokicki THE LEVEL OF LOGISTICS AND THE SOLUTIONS IN THE FIELD OF INFORMATION TECHNOLOGY APPLIED IN AGRIBUSINESS ENTERPRISES... 59

4

5 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) CRIMINAL PROCEDURE MANAGEMENT BASED ON BPM SIMULATION RYSZARD ANTKIEWICZ, MICHAŁ DYK, RAFAŁ KASPRZYK, ANDRZEJ NAJGEBAUER, DARIUSZ PIERZCHAŁA, ZBIGNIEW TARAPATA Military University of Technology (WAT), Cybernetics Faculty The Polish criminal procedure is strictly based on the Polish Constitution and complex legislation. It involves many steps, phases and stages and creates very complex processes that are composed of multiple sub-processes and activities. Nowadays, the participants of the Polish criminal procedure are provided with modern IT limitedly. Increase the effectiveness of criminal justice is possible via changes in both organiational and procedural way, as well as through the IT technologies supporting the processing of information at any stage of a criminal proceeding. We propose two-way analysis of the criminal procedure: modelling and simulation of the process similarly to Business Processes Modelling (BPM), which forms the basis for measuring and assessing an effectiveness, and optimiation based on hybrid models and methods: GAN/GERT with ueuing theory and optimiation. The practical result of this work is the workflow system extended with the business process simulator. Keywords: the criminal procedure, BPM, business process simulation, court performance. Introduction The new century has brought many positive changes stemming from the development of IT technology, the opening of institutions for people and thus simplifying and accelerating many official procedures. However, there are still

6 domains of life in which these changes are taking place very slowly. For proper functioning of a country the third power - the judiciary - is crucial. There is an urgent need for innovation in this area in the three dimensions: organiational, legal (formal) and technology. According to the social research, judicature is criticised by approximately 40% of the Polish citiens [3] in particular, there have been criticied the costs and delays of criminal cases. From the other hand, the public budget allocated to all courts (excluding public prosecution and legal aid), as percentage of GDP per capita, reached 0,38% in 200 (Norway 0,07%, Denmark 0,09%, UK 0,%) [4]. Nowadays, the participants of the Polish criminal procedure (judge, secretary, sites) are provided with modern IT limitedly. Although simple electronic tools for case-management are reported in 50% of courts (mostly civil, commercial or administrative courts, not criminal cases) there are rather stand-alone programs then integrated and distributed multi-function IT systems (e.g. electronic files are available in 0% of courts) [5]. Changes in the criminal process should be preceded by an analysis of the current status and the prospective changes necessary and possible to applying. If we treat criminal procedure in terms of business processes, the selected problems can be solved with the use of tools for process modelling and simulation, which will form the basis for assessing the effectiveness and finally optimiation. Elimination or reduction of bottlenecks in that process, identifying how to redefine and reorganie the process, indicating the time and resource savings is possible to be made after identifying the real causes of problems. Simple examples are the complaints that are allowed by criminal code to submit on almost every stage of the proceedings. Another one is a necessity to service formal defects in criminal process documents (writs, letters, briefs etc.). The subject matter is relevant in studies conducted in the world since the early 990 s. A huge effort has been done to estimate the performance of judicial systems - the analysis of court work from the point of view of economics at [], [2], statistics or prognosis [4] and opinion polls as well as evaluation of the courts by citiens [3]. The paper presents a new approach based upon an example of a district court (case study in the District Court in Bialystok). The introduction briefly describes the background the criminal procedure in Poland. Then we introduce IT tools for analysing the criminal process. Further, a formal model, research goals and the characteristics that describe uantitatively the processes are defined. The main task we have been conducting is a many-components workflow system extended with the business process simulator to analyse and search for the (sub)optimal management of criminal process stages using a computer simulation. Finally, we outline the reuirements and concept of BPM simulator, short case study and conclusions as well. 88

7 2. Business process modelling of the criminal procedure 2.. The background In Poland, there are about 330 courts so-called first instance (district) courts, almost 50 second instance (regional), more than 0 appellate and one supreme (highest) court. In that structure there are ~0,000 professional judges working in courts: ~7,000 at first instance, ~3,000 at second and less than 200 at highest instance [5]. According to general statistics, more than 90% cases are proceeded by courts of first instance. Among different cases a weight of criminal ones is special, so due to the facts the role of criminal departments in district courts is really important. The Polish criminal procedure is strictly based on the Polish Constitution and complex legislation. It involves many steps and stages that at the most general level are: Criminal investigation, Grand Jury proceeding and Execution of sentence. More detailed, there exists the following steps: Reporting of the Crime, Pre-arrest Investigation, Arrest, Arrest and Booking Process, Post-arrest Investigation, Decision to Charge, Filing the Complaint, First Appearance, Preliminary Hearing and/or Grand Jury Proceeding, Arraignment on the Information or Indictment, Pretrial Motions, Plea Bargaining and Guilty Pleas, Trial, Sentencing, Appeals and finally Post-conviction Remedies. The space of steps (and stages of proceeded cases) is perceived as discrete. Depending on goals it can be divided into different number of levels. Actually, this division creates complex processes that are composed of multiple sub-processes and activities. Each one is optional and can run according to various alternative paths thus a procedure can finish without executing all steps. For the proper analysis of a criminal procedure and then the construction of supporting software tools (for workflow management and simulation) there is crucial to adapt a modern approach to business process management (BPM - Business Process Management). It is an approach that allows modern organiations and businesses to operate effectively in the market and respectively to customer needs BPM approach to the criminal procedure The main objective of BPM is to adopt organiations to the needs of their customers. BPM cover the interlinked stages of activity i.e. the identification, analysis & modelling, simulation, implementation, launching and monitoring - which is also the main elements of a typical life cycle of business processes. Assumption is that the processes should be continuously improved on to ensure their accuracy, the ability to better achieve its goals and more competitive 89

8 company. The advantages are many, despite a solid effort reuired to implement and maintain such an approach. According to BPM notation, every step is so called process (at the top level) and subprocess (the subordinated levels), which leads to decomposition to more complex model with more specific tasks. Main area of our interest is judicial criminal proceeding, which is the second step. Figure shows its diagram at a higher level. It begins with the bill of indictment by the prosecutor. The case documents, especially indictment, are then verified by the president of the Criminal Department. This step is a complex subprocess called Indictment verification during which formal verification is performed and reporting judge is chosen. If any formal problems with indictment occurs, subprocess called 20 KPK (20 paragraph of the Code of Criminal Procedure) is started, which means all formal defects should be resolved by prosecutor or indictment will be rejected. After Indictment verification step is completed the reporting judge has all the case files. In the next step, which also is subprocess, these documents are verified by judge and decisions including use of detention for the accused, discontinuance of the proceedings or going for trial are made. Depending on this decisions next steps are trial or session concerning the discontinuance of the proceedings (from which case also can go to trial step or sentencing step). Like steps before, trial is also a complex subprocess with multiple different paths, but unlike them it cannot be supported by BPM system. The last step, which is also subprocess, is sentencing. It is uite complex because events such as appeal or objection can occur. Diagram in Figure may not look as complicated as the judicial criminal proceeding really is. In fact, there are four levels of subprocesses and some of them are executed more than once. For example diagram for process Indictment verification contains fourteen tasks and more than ten alternative paths. What more, two of those tasks are complex subprocesses. One is mentioned earlier, 20 KPK step and the other one is complaint handling process. Both are used many times in further steps of judicial proceeding process. Complaint handling is particularly complex. It contains seventeen tasks and more than twelve different paths and, what causes most problems, loops (decisions made during complaint handling also may be subject to complaint). Considering fact, that almost every decisions made during judicial proceeding may be subject to complaint, it increases complexity of whole process dramatically. Steps responsible for handling defects and complains might be performed in other parts of judicial proceeding - they form so-called pessimistic paths. That is the most complex and time consuming part of this process. One of the main goal of modelling and simulation is to verify what is the impact of pessimistic paths on time of execution and cost of whole judicial proceeding. 90

9 Figure. The higher level of judicial criminal proceeding 9

10 3. Simulation and optimiation for management Criminal procedure life cycle, similarly to BPM, follows the steps outlined at Figure 2. That cycle depends on correctness of modelling and simulation as well as reliable statistics gathered during execution. Thus, there is an obligation to develop adeuate software tools for simulation analysis of effectiveness of the criminal process, which will help finding reasons of problems and allow for proceeding optimiation. Generally, we focus on problems such as congestion, the high cost and delay of procedures elimination or reduction of "bottlenecks" in processes, suggestion of redefinition and reorganiation processes as well as the time and resources saving, etc. Although simulation is very suitable the formal way is adopt hybrid models and methods: GAN/GERT with ueuing theory and optimiation. Figure 2. Model of typical business process life cycle 3.. Formal models and methods To model, optimie and simulate criminal process we can use some formal models and methods well known in operations research. For analysis of complex operations (processes, projects) following models and methods may be considered [8], [9]: ueuing theory, CPM (Critical Path Method), PERT (Program Evaluation and Review Techniue), CPM-cost, PERT-cost, GAN (Generalied Activity Network), GERT (Graphical Evaluation and Review Techniue). In CPM, PERT, 92

11 GAN, GERT models events are described by graph's nodes and activities - by graph's arcs. Each of them has advantages and disadvantages. CPM and PERT methods have limitation to acyclic structure of projects and, in general, they are deterministic. CPM-cost and PERT-cost have extended properties because we may take into account not only realiation time for activities but also cost, but they have the same disadvantages like pure CPM and PERT. Generalied Activity Network (GAN) proposed by Eisner [6] allows us to build project network in which realiation of any event causes start of realiation one of activities, not all of them for the event. GERT network [0], [] is a type of GAN network. In GAN network activities may be deterministic or stochastic; dependencies between events and activities are being done using some logic operations which are described on "enters" to a node and "exits" from a node (see [6]). For "enter" to nodes (events) we have following logic operations: - conjunctive and realiation of the event takes place if and only if all activities which take end in the event are realied; - alternative or realiation of any activity which take end in the event causes realiation of the event; - disjunctive xor (excluding-or) realiation of any activity which take end in the event causes realiation of the event and one activity only which take end in the event can be realied at the same time. For "exit" from nodes (events) we have following logic operations: - conjunctive and realiation of the event causes realiation of all activities which take start from this event; - disjunctive xor (excluding-or) if the event is realied, then one and only one activity which take end in the event is being realied with fixed probability at the same time; sum of realiation probabilities of all activities which take start in the event provided that the event was realied is eual one. In the case when all enters and exits for all nodes are conjunctive ("and"- "and") then the network is deterministic (CPM, PERT). Otherwise, the network is stochastic (GERT, GAN). Let's notice that in the crime process at the start and the end of any node in network we can have each of situations presented above. In Figure 3 we describe, as an example, a part of crime process deals with the first steps after submission of claim form. Figure 3. Example of GERT (GAN) analysis in crime process Source: on the basis of [8] 93

12 Activities (arcs) in the network have following interpretation: - a submission of claim form (for the first time); - b formal control of claim form; - c transfer of claim form for further processing; - d submission of claim form after removing formal defects; - e euivalent of path from node 2 to node 4 taking into account cycle 2-b-3-d-2. Events (nodes), 2, 3 and 4 describes, respectively: start of the process, event directly after submission claim form, event directly after formal control of claim form and event directly after transferring of claim form for further processing. Nodes 2 and 3 have alternative "enter" and "exit", node has conjunctive "enter" and "alternative" "exit" and node 4 has alternative "enter" and conjunctive "exit" (in this case node 3 should has disjunctive "exit" and node 2 - disjunctive "enter" but for the example of using Elmaghraby's graph algebra the assumption is enough). It causes that the network is stochastic. Let's take into account that the network has the cycle 2-b-3-d-2. This fact causes that we can't use classical methods like CPM or PERT because in these methods it is reuired the network (graph) to be acyclic. Taking into account Elmaghraby's algebra [7] for GAN/GERT networks we obtain following realiation probability and time of the "e" activity [8]: n+ n a b c e = a ( b ) ( d ) c = n= 0 Pb Pd P P P P P n+ n [ ( n ) n ] P ( P ) ( P ) P () e a b c d a b d c n= 0 τ a + τ b + τ c τ d + Pb Pd + Pb Pd Pb Pd = Pa Pb Pc P P b d P P P τ = τ + + τ + τ + τ = where: P x - realiation probability of the x activity; τ x - realiation time of the x activity; n - number (count) of repetitions of the cycle 2-b-3-d-2. A case when a cycle in a network has more than 2 arcs can be reduced to a case with two arcs only (like in the cycle 2-b-3-d-2) and we can use the procedure described by euation (). Conclusions: (a) PERT and CPM (PERT-cost, CPM-cost) - graph must be acyclic; only "and" enters to nodes and exits from nodes can be modelled; deterministic graphs (even in PERT because of deterministic euivalent of characteristics described on nodes and arcs); 94

13 (b) GAN, GERT - "stochastic" events and activities give possibilities to model wider class of problems; possibility of modelling not only "and" enters to nodes and exits from nodes but also "or" and "xor"; problems to set probabilistic distribution of characteristics for events and activities The simulation component description A complementary approach is BPM simulations where the basic expectations regarding proposed software tools concern the ability to construct the model of the criminal process, i.e. identification of this process and its editing (phases, activities, events, sent documents and the relationship between the components of the process), and then verify the effectiveness using simulation, including estimation of the value of certain parameters characteriing the process, e.g. the time reuired for the process or a fragment thereof, or the number of necessary resources needed for its running. We want to verify what is the impact of pessimistic paths on time of execution and cost of whole judicial proceeding. Identification of the parameters (potential criteria for assessment) the so-called KPIs (Key Performance Identifiers) to assess the effectiveness of the process should be conducted mostly based on the practical experience of law enforcement, prosecution and the judiciary. This could be done by making a specialied survey among experienced investigators, prosecutors and judges. Nonetheless, based on analyses carried out under this project, a comprehensive list of criteria for the evaluation process can be proposed. It is also expected that constructed software tools will enable optimiation of the criminal process involving e.g. searching "bottlenecks" of the process, an automatic (semi-automatic) reorganiation of the process to achieve extreme (min or max) or reference value of the specified parameter/s of the criminal process. An important feature of the constructed software tools should be an intuitive graphical user interface (in particular for the domain expert, such as a policeman, prosecutor, judge). For this purpose, it is proposed to use and adapt BPMN (Business Process Modelling Notation) to the specificities of the criminal process. The proposal to use BPMN follows that the notation has become a global standard for modelling business processes for different organiations of any structure and any level of complexity of the processes occurring in them. An important advantage of BPMN is the simplicity of implementation and deployment (automated transformation into an executable language such as BPEL4WS Business Process Execution Language for Web Services) of the modelled processes within an organiation, together with the support of IT systems, thanks to BPMN compatibility with concept of SOA (Service Oriented Architecture). 95

14 Figure 4. The main use cases for the IT system The main component (Figure 5) is SimulationEngine consisting of JGraph and Deskit, which are together responsible for simulation of the judicial proceeding process. We have relied on mixed discrete simulation concept: event-driven and process-oriented. As the central elements of the simulation they provide data for the other components (by providing services for them), which are the Optimier and Statistics. Data for visualiation provides an interface called ProcessView. Depending on reuirement, it may be used for standalone applications or web applications. Simulation, optimiation and calculation of statistics reuires a process model and setup of the experiment. They are provided by two components: ProcessModels and Configurator. The last one is also responsible for the conversion from JPDL or BPMN2 to graph model handled by simulation tool. 96

15 Figure 5. The architecture of simulation components 97

16 4. Conclusion General goal of the conducted project is elimination all reasons that might weaken the access of citiens to judiciary system, as well as make hard to keep the guarantee of euality before the law. To date, the greatest efforts for accelerating criminal proceedings and limiting delays of cases has been directed to increasing the budget of judicial systems it has been noted in Poland and also Belgium, Luxembourg, Austria [5]. Such a strategy is proper only for short or transitional time. There is high time to apply other methods based on IT cutting-edge tools and operational research solutions. Therefore, we propose two-way analysis of the criminal procedure: modelling and simulation of the process similarly to Business Processes Modelling (BPM) and optimiation based on hybrid models and methods: GAN/GERT with ueuing theory and optimiation. In the case of optimiation of single subprocess (inside whole crime process) we obtain locally (for single subprocess) optimal solutions. In such organiations like judicature (but not only) we must focus on optimiing the operation of whole organiation (globally optimal solutions). It seems to be essential to use hybrid models and methods: GAN/GERT with ueuing theory and optimiation of resource (people, courtrooms, time,...) allocations [9]. The results of this work is the workflow system extended with the business process simulator case study is being organied in the District Court in Bialystok. Although the proposed approach is a practical step on the way to alteration of criminal judiciary, some expected benefits are strictly determined by amendments in the law (including Code of Criminal Procedure) keeping up innovative capabilities acuired with IT systems. Acknowledgement The results are applicable to the project No O ROB 002 0/ID 2/2 co-financed by NCBiR (The National Centre for Research and Development). REFERENCES [] Rosales-Lope V. (2008) Economics of court performance: an empirical analysis, European Journal of Law and Economics, Vol. 25:23 25, Springer, New York / Heidelberg. [2] Posner R. A. (2005) Judicial behaviour and performance: an economic approach, Florida State University Law Review, Vol. 32: , Florida State University, USA. 98

17 [3] Beldowski J., Cikowic M., Sescilo D. (200) Efficiency of Polish judiciary from the point of view international and national research (Polish language edition), Helsinki Foundation for Human Rights, Civil Development Forum, Warsaw. [4] European Commission for the Efficiency of Justice (202) European judicial systems - Edition 202. Efficiency and uality of justice, ISBN , Strasbourg. [5] Dubois E., Schurrer C., Velicogna M. (203) The functioning of judicial systems and the situation of the economy in the European Union Member States, Report prepared for the European Commission (Directorate General JUSTICE), Strasbourg. [6] Eisner H. (962) A Generalied Network Approach to the Planning and Scheduling of a Research Project, Operational Research, Vol. 0 (), [7] Elmaghraby S.E. (964) Algebra for the Analysis of Generalied Activity Networks, Management Science, 0, pp [8] Ignasiak E. (982) Teoria grafów i planowanie sieciowe, PWE, Warsawa. [9] Laslo Z., Golenko-Ginburg D., Gonik A. (2005) Alternative stochastic network projects with renewable resources, Computer Modelling and New Technologies, Vol. 9-, pp [0] Pristker A.A.B. and Happ W.W. (966) GERT: graphical evaluation and review techniues, Part I (Fundamentals), Journal of Industrial Engineering, 7 (5), pp [] Pristker A.A.B., Whitehouse G.E. (966) GERT: graphical evaluation and review techniues, Part II (Probabilistic and industrial engineering applications), Journal of Industrial Engineering, 7(6), pp

18 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) 00 2 GAME-TREE STRUCTURES WITH THE COMPLEX COMPLEXITY LEVEL AS A TOOL IN KNOWLEDGE ENGINEERING ADAM DEPTUŁA Department of Knowledge Engineering, Opole University of Technology The paper concerns application of the dependence graphs and game- tree structures with the complex complexity level as a tool in knowledge engineering. Different graph solutions mean connections between input and output data as well as decision variables of the analyed system. The graph distribution from any vertex in the first stage leads to a tree structure with cycles, and next to a general tree game structure. Game-tree structures from each node have different shape and properties. Algorithmic way to create graphical structures out of a mathematical model describes the optimiation method of systematic exploration. Tree structure, with the lowest values of complexity level is the simplest structure. Keywords: knowledge engineering, dependence graphs, game- tree structures, decision- making, artificial intelligence. Introduction The subject of knowledge engineering and expert systems is considered to be one of the most important areas and directions of contemporary IT development. The main task of the widely understood knowledge engineering is to gather and formalie experts knowledge into the form of rules used by expert systems and decision support systems. Expert systems, as IT systems, are aimed mainly at solving specialist problems reuiring professional expert s opinion based on knowledge so they are connected with gaining and processing it. The main idea of

19 an expert system is the expert s knowledge transfer to the system euipped with the knowledge base. The knowledge representation is a basic term for different kinds of decision processes and drawing conclusions and it is one of major problems which has not been completely solved yet. In the knowledge engineering, knowledge representation is treated as a way of presenting, in the formal language, the whole knowledge range reuired for intelligent system maintenance [, 2]. However, their main task is to support the decision making process on the basis of the possessed knowledge. Decision making processes place an emphasis on modern information processing techniues in order to improve the decision making process. Both in the knowledge creation for the enterprise needs and in specifying construction guidelines for the designer, it is necessary to use an appropriate device describing the decision making process. First decisions are made in the planning process and then, they are reinforced are corrected. There is a wide range of tests concerning working out of a methodology supporting decision making processes and controlling in the systems with different complexity scale with the participation of the artificial intelligence [9, 0]. Among tools supporting decision making processes, it is possible to enumerate: decision tables, dendrites, decision logical trees, tree classifiers as well as logical comparisons and graphs. This article presents dependency graphs concerning signal transmission and game structures as tools supporting decision making processes in the knowledge engineering. The mathematical model of the analysed system (element) is a group of functions which are connected with one another by means of different variables and in this way, they describe connections between figures in the system. As a result of saving and making a decomposition of a graph presenting these functions dependencies, we obtain decomposition groups which structurally describe properties of subseuent sub-systems of a given system. The obtained game structures differ in shape and properties. In order to choose a game structure with the lowest complexity, it is necessary to calculate a complex complexity coefficient for all structures that are obtained. 2. The dependence graph for tree game structures Directed graph defined as an ordered pair of sets. The first of these contains vertices and the second consists of the edges of the graph. Figure shows an exemplary dependence digraph of the signal flow. 0

20 Figure. Dependence digraph of the signal flow Source: Own study based on [3, ] Depending directed graph of Figure consists of a set of vertices Q : Q = {,, } and edges from the set: Z {, } 2 3 =. The graphs distribution from any vertex in the first stage leads to a tree structure with cycles, and next to a general tree game structure. Each structure has a proper analytic notation G + and G ++ i. The algorithm describes the distribution of dependency graph is presented in [5, ]. Decomposition of the dependence graph of the signal flow from the defined initial vertex G + in the first stage leads to the expression: i G = ( ( 3 (, 2 3 ), 2 2 ( 3, 2 3 ) ) ) and then to express G ++ : i G = ( ( 3 (, 2 3 ), 2 2 ( 3 (, 2 3 ), 2 3 (, 2 3 ) ) ) ) where: i - the i-th edge r extending out from the r-th vertex i. r Considering tree structures we must determine element subordination in the system. Each structure has its proper analytic notation ( G + i and G ++ i, where i is the vertex from which the graph decomposition was started). Each element r has always subordinated elements i. Both r and i elements can occur many times in the expression G ++ k i in the brackets (...) k with different values of k, i.e. at various stages of the tree structure [5, 6, ]. Figure 2 shows the tree structure with cycles, while in Figure 3 a tree game structure from the initial vertex. i 02

21 Figure 2. Tree structure with cycles and the initial vertex Source: Own study based on [5, ] Figure 3. Tree game structure from the initial vertex Source: Own study based on [5, ] 03

22 3. Complex complexity coefficient for game-tree structures The level of structure's complexity of is determined by the complex complexity coefficient L( G ++ ) [3, 4, 8]: i + + d ( wi ) L( Gi ) = () h( w ) + w W ( L) where: L( G ++ ) - complexity coefficient of structure i G ++, i wi - i-th node, d( wi ) = deg( wi ) - degree of i-th node branching (amount of node branchings), h( w i ) - distance from the i-th node root, W ( L) - set of all nodes. In order to state the importance of game structures obtained as a result of a dependency graph decomposition from each graph vertex, a complex complexity coefficient is introduced [8] L K ( G ++ ): i i K ++ d( wi ) L ( Gi ) = + h( w ) + w W ( L) where: L K ( G ++ )- comoplex complexity coefficient of structure i G ++, i L- number of leaves for the i-th node branching ( deg( wi ) 2 ), hl i - amount (complexity) of the i-th leaf. Then we obtain a better uality evaluation of a given structure by means of taking into consideration decisions which are not branched. The current structure complexity coefficient defined the structure in the macro scale with taking into consideration only the amount of nodes in the structure and the degree of their branching. The complex complexity coefficient of a structure takes into consideration the complexity degree of all leaves going out of each branching node [8]. The Figure 4 shows an example game-tree structure with different coefficients L and L K. i l L L h li (2) 04

23 Figure 4. Tree-game structures with different complexity coefficients L and L K Source: own study based on [3, 8] 4. Game graphs with different vertex connections as tool in knowledge engineering An analysis of a mathematical model of a given system can be described by means of game graphs with different exemplary vertex connections G(), G(2) presented in the Figure 5. Figure 5. Dependence digraphs G(), G(2) for different connections vertex For a graph G() as shown in Figure there is a set D() tree-game structures: { } D() = G(), G(), G(), G(), G() Tree-game structures are shown in figures

24 G ( ) : G ( ) : Figure 6. Tree game structures G () ++ i G () G () : G () : G () : Figure 7. Tree game structures G () ++ 3, G () ++ 4 i G ()

25 Analytic expressions G () + i and G () ++ i for the structures of the figures 6-7: G() + = ( ( ( (, ( ( ) ), ) ) ) ), G() ( ( ( (, ( ( ) ), ( ( ) ) ) ) ) ), = G() = ( ( ( ( ) ), ( ( ) ), ) ), G ( ) ) ) ), ( ( ( ( ) ) ) ) ) ), () + 2 = ( 2 ( 4 ( 2 3 ( 3 2 ) ), 5 4 ( 5 ( 7 ( G() + = ( ( ( ( ), ( ( ) ), ) ) ), G ( ) ) ) ) ) ), () ++ 3 = ( 3( 3 2 ( 4 ( 2 3), 5 4( 5( 7 ( 2 3 ) ) ), 6 4( 5( () ++ 4 = ( 4 ( 5( 7 ( 2 3( 3 2 ( 4, 5 4, 6 4 ) ) ) ) ) ), G G () ( ( ( ( ( (,, ) ) ) ) ) ), = G () ( ( ( ( (, ( ), ( ) ) ) ) ) ) = For each of the structures G () ++, G () ++ 2, G () ++ 3, G () ++ 4 i G () ++ 5 was calculated complex complexity coefficients by the formula (2): K d( wi ) L L ( G() ) = + = 0, = w W ( L) h( wi ) l L h + + l i L K ( G () ++ 2 ) = 8,66, L K ( G () ++ 3 ) = 8,66, L K ( G () ++ K 4 ) = 3,33, L ( G () ++ 5 ) = 4,0 For a graph G(2) as shown in Figure there is a set D(2) tree-game structures: { } D(2) = G(2), G(2), G(2), G(2), G(2) Tree-game structures are shown in Figures

26 08 Figure 8. Tree game structures (2) G ++ i 2 (2) G ++ Knowledge engineering is an interdisciplinary branch of solving many different problems as far as exact science and sociology are concerned. Research issues include among others: artificial intelligence, expert systems and in particular, decision support systems. In any of the analysed systems, in the newly recommended work conditions, unknown dependency functions depending on time will change their behaviours. This is why, one rule is in force: graph vertices (the Q set) define functions depending on time, whereas edges (the Z set) define decisions. Game tree-structures from each graph vertex describe the decision making process and the space of the possible to get states of the analysed system

27 4 5 8 G (2) : G (2 ) : G (2) : Figure 9. Tree game structures G (2) ++ 3, G (2) ++ 4 i G (2) Dependency graph in the description of a mathematical model of a machine set When we use a dependency graph (Figure ) in the description of a mathematical model of a machine set (Figure 0), which is described by means of a set of algebraic, differential and integral euations, then the graph vertex is formed by evaluated functions depending on time, whereas decisions are formed by construction and / or exploitation parameters as well as analytic and algebraic transformations [5, 6, 7]. 09

28 Figure 0. Hydraulic scheme of the system Source: own study based on [5, 6, ] Figure. Dependence digraph of the signal flow G(H) for the hydraulic system Source: own study based on [5, 6, ] For a graph for hydraulic system G(H) as shown in Figure there is a set D(H) tree-game structures: D( H ) = G( H ), G( H ), G( H ), G( H ), G( H ) { Qs Ps ω Pp Qp} The Figure 2 shows the game-tree structures G ++ ω and G ++ Qs from the initial K vertex ω and Qs with complex complexity coefficients L ( G( H) ++ ω ) = 7,0 and L K ( G( H ) ++ ) = 8,66. Qs 0

29 ++ Figure 2. Tree game structures G( H) ω i G( H ) ++ Qs Source: own study based on [5, 6, ] 6. Conclusion Game graphs make it possible to analyse the so-called connected decisions. Results obtained after the first decision have an influence on subseuent decisions. This is why they make it possible to make dynamic models. The complexity coefficient makes it possible to choose a structure of the smallest decision making degree. There are many methods of looking for ideas and problem structuration in the knowledge engineering e.g. a morphological method, an analysis of connected AIDA decision areas, a decision tables method, etc. Game graphs are among these tools. It is worth highlighting that there is a dependency between the complexity coefficient value and the sie of the multi-valued logical decision tree.

30 REFERENCES [] Allard C. R., Ouwersloot H., Lemmink, J. (2006) Antecedents Of Effective Decision- Making: A Cognitive Approach, The IUP Journal of Managerial Economics, IUP Publications, vol. 0(4), pp [2] Brunsson N. (2007) The Conseuences of Decision-making, New York: Oxford University Press. [3] Deptuła A. (20) Determination of game-tree structures complexity level in discrete optimiation of machine systems, in: Proceedings of International Masaryk Conference for Ph. D. Students and Young Researches, Hradec Kralove, Cech Republic, pp [4] Deptuła A. (203) Kompleksowy współcynnik rogrywający parametrycnie grafu ależności prepływu sygnałów, XLII Konf. Zast. Mat., Zakopane 203, Inst. Mat. PAN, Warsawa. [5] Deptuła A., Partyka M.A. (20) Application of dependence graphs and game trees for decision decomposition for machine systems, Journal of Automation, Mobile Robotics & Intelligent Systems, vol. 5, No. 3, pp.7-26 [6] Deptuła A., Partyka M.A. (200) Application of game graphs in optimiation of dynamic system structures, International Journal of Applied Mechanics and Engineering, vol.5, No.3, pp [7] Deptuła A., Partyka M.A. (203) Discrete optimiation of a gear pump after tooth undercautting by means of complex multi-valude logic trees. XVI Konferencja Innowacje w Zarądaniu i Inżynierii Produkcji, Zakopane 203, ;Pol. Towar. Zar. Prod. PTZP. (in Polish) [8] Deptuła A. (202) Współcynnik łożoności struktury dla minimaliacji wielowartościowych funkcji logicnych; XLI Konf. Zast. Mat., Zakopane 202, Inst. Mat. PAN, Warsawa. [9] Hall J.G. (202) Engineering knowledge engineering, Expert Systems, Volume 29, Issue 5, pp [0] [] Kaimiercak J. (978) System cybernetycny, Wieda Powsechna, Omega, Warsawa. 2

31 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) 3 25 APPLICATION SUPPORTING COMMERCIAL BANK S CAPITAL MANAGEMENT VIERA GAFRIKOVA a), ZDZISŁAW ODRZYGÓŹDŹ b), WIESŁAW SZCZESNY c) a) Faculty of Management and Finance, Chodkowska University b) Faculty of Mathematics and Information Science, Warsaw University of Technology c) Faculty of Applied Informatics and Mathematics, Warsaw University of Life Sciences This paper describes the basis - resulting among others from the prudential capital regulations - for capital management at a commercial bank. In the capital management process there are used capital allocation and risk adjusted performance measures. Based on this, the concept of the central MIS extension has been presented in a way that allows to include a wide range of different level managers in the capital management process in order to improve the capital efficiency from the perspective of shareholders. Keywords: Capital Management, Capital Allocation, Risk Adjusted Performance Measures, Business Intelligence, Management Information System. Introduction The bank s capital, like capital of any other enterprise, constitutes paid in capital contributed by owners and earned capital that comes from profits of the company in the course of business. It is created in accordance with applicable laws, regulations and statute of the bank. Commercial bank capital fulfills many important functions, although its share in total liabilities is relatively low (for example, in the banking sector in Poland it amounted to 0.4% as of the end of June 203 []). The basic function of capital is the buffer function that ensures the bank s survival in case of unexpected losses incurred.

32 The capital can be viewed from several perspectives. From a purely accounting point of view capital is treated as a free source of funding, primarily for fixed assets and investments in risk-free instruments (for example treasury bonds). Shareholders and their representatives (the Supervisory Board and the General Meeting of Shareholders) expect an adeuate return on capital, e.g. in the form of dividends paid. This means that the bank is expected to generate profit depending on the sie of capital and cost of capital in the market. Thus, the higher is the bank capital, the higher profit should be generated to share the profit with shareholders, therefore, from this perspective, capital can be associated as the cost. From the perspective of stakeholders such as creditors or institutions that represent them (regulators and rating agencies) capital is seen as a factor determining the sie and potential of the bank and an additional security buffer. The perception of the bank affect both the number and type of bank customers. This causes that the problem of optimal use of capital held - in the times of capital deficit - has become one of the most important issues that should be solved with the active participation of a large group of employees. Efficiency in the use of capital can be improved already starting with activities of linear employees. Therefore, in the majority of large banks dedicated modules (applications) are created within the central MIS, that support the absorbed capital efficiency management process, involving fairly large group of employees. The purpose of this paper is to present the concept how to incorporate the application concerning efficiency management of capital absorbed by business units into the standard central Management Information System (MIS). From the bank s management point of view, the role of this module is to support the optimiation of the capital sie and structure and capital profitability (a compromise among the return for shareholders, regulatory reuirements, rating agencies criteria and cost of financing). However, it should be noted that in order to implement the absorbed capital efficiency management process by business units effectively, the system of cost management, including the cost allocation to business units, and system of management by objectives should work at the bank. The implementation of the management by objectives concept on the basis of the central MIS is described in [2], [3]. When designing the application, the number of reuirements set by supervisors should be taken into account. These reuirements are discussed in the next section. 2. Supervisory capital reuirements 2.. Basel Agreements The activities of banks are subject to extensive regulations, including reuirements related to capital. In accordance with the Polish Banking Act "In order 4

33 to ensure their economic safety, banks shall be reuired to possess own funds adjusted to the scale of the operations they conduct (art. 26, [4]). External rules reuire banks to maintain a minimum level of capital, which reduces the probability of their failure, and which in turn increases the stability of the entire banking system. The capital reuirements imposed on banks are based on international standards set by the Basel Committee on Banking Supervision, which has developed a set of recommended practices for financial risk management in the banking sector, security and the level of capital banks must hold. Recent changes in the prudential recommendations, involving the introduction of stricter reuirements on bank capital adeuacy and liuidity, were forced by the financial crisis that swept through financial markets in the years Changes are described in the document known collouially as Basel III, published in 200 [5]. Prudential standards based on Basel III were transposed to the European legal order in the form of two acts: Regulations [6] and the Directive [7] of the European Parliament and of the Council published in June 203. The new standards will be effective from January 204 with a number of transitional periods Capital Adeuacy Assessment Banks are reuired to assess their capital adeuacy in two areas: within the socalled Pillar I and Pillar II. Pillar I precisely defines risks that must be taken into account when assessing the capital adeuacy as well as methods of capital reuirements calculation for individual types of risks: credit, market, settlement, delivery and counterparty credit risk, operational risk. The capital reuirements under Pillar I are called regulatory capital reuirements. Pillar I capital adeuacy is determined in relation to the regulatory capital (see section 2.3.), i.e. the regulatory capital may not be less than the total capital reuirements. Under Pillar II, banks are reuired to estimate the amount of (internal capital) needed to cover all of the material risks in the bank's operations and changes in the economic environment, taking into account the expected level of risk. Risk measurement and determining the resulting capital reuirements are generally performed using models of Value at Risk (VaR). Pillar II capital adeuacy assessment is based on comparing the internal capital with so-called available financial resources (AFR). Definition of AFR depends on supervisory regulations in the given country: in some countries AFR are set eual to the regulatory capital, and in other they may differ from the regulatory capital Regulatory capital The regulatory capital definition is based on the definition of the accounting capital (that is the capital recorded in the bank s balance sheet within the meaning 5

34 of the applicable accounting framework). It is, on the one hand, broader than the accounting capital definition, as it may include additional items, which are not included in the accounting capital, and on the other hand is narrower as the different kinds of deductions are taken into consideration in the regulatory capital. The regulatory capital is composed of two main categories of capital: Tier capital and Tier 2 capital. Tier capital consists of common euity Tier capital (or core Tier capital) and additional Tier capital. Common euity Tier capital is the highest uality capital and includes funds contributed by the owners and items created from profits earned. At the commercial bank it includes - among others - ordinary shares or other financial instruments, issued directly by the bank, and that meet a number of strict conditions, reserves, retained earnings, funds for general banking risk, interim profits (under certain conditions). Among deductions applicable to common euity Tier capital one can uote items such as intangible assets or significant investment in financial sector entities. The Additional Tier capital includes for example financial instruments, not necessarily issued by a bank that meet a number of conditions, generally less stringent than the conditions for the classification of instruments as Common Euity Tier capital. The Tier 2 capital, includes, inter alia, subordinated loans fulfilling a number of strict conditions, such as subordinated loans have an original maturity of at least five years and the nominal amount of loans in the final five year period of their contractual maturity decreases linearly (proportionally) to the time left to maturity and the claim on the principal amount of is wholly subordinated to claims of all non- subordinated creditors Minimum capital levels Under Pillar I, capital adeuacy ratio (CAR) is used to measure the adeuacy of the bank's capital base. This indicator is defined as the ratio of capital to 2,5 times the total capital reuirements. Banks are obliged to maintain a capital adeuacy ratio at the level not lower than 8%. This means euivalently that bank s capital must be greater than total capital reuirements. In accordance with the regulations [6], banks are obliged to calculate besides CAR also Core Tier (CT) ratio and Tier ratio (T). They are defined as the ratio of Core Tier capital /Tier capital/ to 2,5 times the total capital reuirement. The lower limit of CT ratio was set at 4,5%, and of T ratio at 6%. In addition to minimum capital reuirements described above, banks are reuired to maintain - mandatory or optional extra capital, so-called capital buffers referred to the Core Tier capital, which in practice means that CT ratio has to be higher than 4,5%. 6

35 Introduction of three types of capital adeuacy ratios and minimum limits for them intends to force banks not only to maintain capital at an appropriate level, but also to provide the proper uality of the capital structure The determinants of capital held by banks Banks generally maintain capital at a higher level than is reuired by law, for example, the average capital adeuacy ratio for the banking sector in Poland amounted to 4,7% at the end of 202 [8]. Supervisory reuirements are not the only factors that affect the amount of capital held by banks. Among other factors one can name: an internal risk assessment performed by the banks themselves depositors expectations in view of safety reasons the expectations of shareholders due to the reuired return on capital planned development of the bank s activities criteria of credit rating agencies future regulatory changes in the capital reuirements - this is the current case, when banks are preparing for the implementation of Basel III recommendation of banking supervision - for example, the Polish Financial Supervision Authority concluded that the dividend for the year 202 might be paid only by banks that had met several conditions, including capital adeuacy ratio above 2% and the Tier capital ratio above 9%. 3. Capital management 3.. Main areas of capital management The capital management process consists of two main areas: the capital adeuacy management and capital allocation. Capital adeuacy management is closely related to the identification, measurement and monitoring of material risks to which banking activities are exposed. The main objectives of capital adeuacy management are: ensuring the safe running of the bank in normal and extreme conditions - by maintaining a balance between the ability to take risks (limited by capital owned) and the level of risk generated to maintain the capital covering risks at an optimal level for the bank, above the statutory minimum, enabling further development of activities and protection of shareholders interests to keep the preferred capital structure in order to maintain the desired uality of capital to cover risks 7

36 The capital allocation means the assignment of the appropriate part of the bank s capital down to lower business units such as organiational units, transactions, business lines, products, customers etc., in order to be able to create value for shareholders by maximiing the return on conducting business activities, taking into account the risk appetite (see e.g. [9], [0]). The capital allocation allows the calculation of return on capital for different banking activities and an evaluation of managers responsible for given activities on the basis of the allocated capital. Dedicated module / application in the central Management Information System can serve to achieve these goals Risk adjusted performance measures Risk adjusted performance measures (RAPM) can serve as the key to a wellorganised, well-structured and transparent capital management process. RAPM compare financial results, for example profits from a particular activity, to the capital amount reuired to produce these results, taking into account the risk involved in the activity. The traditional measure of return on capital (ROC or ROE Return on euity) is defined as net income generated during the period divided by average shareholder's euity for the period. This is a very general definition, and in practice there are wide variety of approaches to measure both the net income and the capital. Adjusting by risk the numerator or the denominator (or both) we obtain risk adjusted performance measures. The most commonly used indicators are: RORAC (return on risk adjusted capital), RAROC (risk adjusted return on capital) and RA- RORAC (risk adjusted return on risk adjusted capital). All these indicators are generally calculated taking into account cost of capital (the capital is not treated as the free source of financing). Capital definitions used in the RAPM methodology generally are closely related to the capital definitions used in the capital adeuacy measurement. The formula () presents one of possible ways to calculate RARO- RAC indicator: EVA RARORAC = () AC where EVA means Economic Value Added, also referred to as economic profit and AC capital at risk. The amount of capital at risk depends on the bank s risk appetite and the target level of capitaliation. We distinguish between two types of capital at risk: capital absorbed to cover the risk that has already been undertaken (calculated during the monitoring process) and allocated capital to cover the risk in the future (calculated in the budgeting and planning process). Allocated /absorbed/capital is calculated on the base of regulatory capital reuirements (RC) and the internal capital (IC), namely the greater amount of these two is taken as allocated capital: 8

37 9 ( RC IC) AC = max, Capital RC is calculated according to the formula (2): RC = CTratio.2,5. REG (2) where CTratio target value of the Core Tier ratio established by a bank see subsections 2.4., 2.5, REG total capital reuirements for all regulatory risks see subsection 2.2. If the available Core Tier capital of the bank (DC) is higher than the allocated capital (AC), the bank deals with a capital surplus, otherwise (DC <AC) with a shortage of capital. To calculate EVA, the formula (3) is used: EVA = NOPAT C. R i (3) where NOPAT means net operating profit after taxes, C invested capital, R i - unit cost of capital. The invested capital definition depends on the methodology used by a bank - for example, it can be treated as the allocated capital increased by regulatory deductions from CT capital. If the EVA is positive, the company creates value for shareholders, negative EVA indicates that shareholder value is destroyed. The cost of capital is usually defined in line with the methodology CAPM - Capital Asset Pricing Model (see for example [9], p. -4). In this model, it is assumed that the reuired return on investments is determined on the basis of historical data according to the formula (4): Ri = R f + β. Erp (4) where R f risk-free rate (usually the return rate on treasury bill or treasury bonds), Erp -euity risk premium is the excess market return over the risk free rate, β the proportionality factor (sensitivity to market risk). The NOPAT can include various types of adjustments of the net profit, depending on the bank s methodology. The most typical is the exclusion of the result on extraordinary items after taxes. Sometimes so called virtual interest (VI) calculated on a surplus or on a shortage of capital are taken into account in the NOPAT calculation: NOPAT = NP EI VI ( TR) where NP net profit, EI result on extraordinary items after taxes, VI = R f (AC-DC) virtual interest, R f risk free rate, TR tax rate. Virtual interest increase the net profit, in the case of a capital shortage, and decrease the net profit in the case of a capital surplus. This allows the better comparison of different entities with different amounts of capital, as the entity with greater amount of capital is supposed to generate profits easier.

38 3.3. Capital allocation to transactions and clients In order to include managers with different levels of responsibility in the process of absorbed capital efficiency management, capital should be allocated to the level of single transaction completion of which reuires the bank to hold capital. Only reports which include amount and cost of assigned capital and the full allocation of administrative costs show to business units full financial impact of the transaction. Therefore, the report on the EVA at the client level shows full information on effects of customer-bank relationship during the reporting period. Usually, the allocation methodology is based on the technical capabilities of the Management Information System implemented in the bank. Currently, even the simplest MIS includes a module of Funds Transfer Pricing (FTP) which allows to calculate net interest margin on a product-by-product basis (or on other lower than bank level in which we are interested in). Figure shows a scheme of how to extend the valuation of the typical credit transaction which is valuated in MIS by means of FTP and reuired 8% of capital. 00 PLN LOAN 8 PLN BONDS. Funding 2. Capital raising 3. Placement capital 00 PLN Financing 8 PLN CAPITAL Figure. Example of assigning the capital to credit transaction (assuming the reuired capital at 8% level) illustrating fund flows used in the valuation of the funds according to bank s FTP system From Figure follows that, from the business unit point of view, capital allocation causes the additional charge for the transaction net interest income, because interest revenues from assets at 8 PLN are lower than cost of capital at 8 PLN. A similar idea can be used in the process of allocating capital to other types of transactions. 4. Construction of capital management system Due to external regulations, more or less developed capital adeuacy management process is established in each bank. However, there is lack of available 20

39 analyses and studies concerning the degree of use of the capital allocation process for the capital management in banks. Some banks, including the two biggest Polish banks, PKO BP (see [], p.0) and Pekao SA (see [2], p. 50), put general information on the functioning of their capital allocation processes in their financial statements. A little bit more information can be found at specialied conferences, for example SAS Forum [3]. Detailed solutions, however, constitute the business secret of banks and are unknown to others. To fulfill the obligations towards supervisors there is no need to build a very extensive support system taking into account the broad participation of business units. However, the inclusion of a broad group of participants enables significant increase of the absorbed capital economic efficiency. When building a capital allocation system in banks there should be taken into account the following elements: selection of an appropriate methodology (the conseuences of various performance measurement models must be thoroughly thought and models should be detailed described) establishment of standards for specifying the capital to be allocated and algorithms for capital allocation to lower levels (e.g. activities, transactions) definition of NOPAT calculation algorithm at the whole bank level and at the lower levels. It should be decided what kind of costs should be considered at the lower levels: the direct costs only or also indirect costs? In the case of indirect costs the allocation methods for these costs are needed. The net profit includes the various cost items, for which the lower level managers are not responsible, in particular for general administrative costs or taxes to impose these costs on managers may be demotivating for them, establishment of centres of responsibility for capital management results, establishment of rules for reporting the financial results taking into consideration cost of capital employed When building the supporting system for the capital management at a bank, we must keep in mind that we are dealing with an integrated process that is based on supervisory regulations and internal procedures approved by the Supervisory Board and the Management Board. Therefore, a solution should be flexible, because existing procedures in this area are subject to periodic review to reflect changing internal and external conditions of the bank s activities. For this reason, the simplest solution seems to be to include to the central MIS the capital management center and based on its functions to modify the internal funds flow and their pricing within the current Funds Transfer Pricing system. This approach allows a fairly easy introduction of a passive capital allocation to the lower levels (for example clients) and after the implementation of the EVA /RARORAC/ reporting at the level of the individual business units to go into more difficult phase of the use of EVA /RARORAC/, namely using them as a Key Performance Indicator (KPI) in the system Management by Objectives for the responsible managers that enables to 2

40 increase gradually the efficiency of capital absorbed by individual business divisions. Support can be organied using a dedicated application that uses data from systems supporting the risk management and the detailed information from financial accounting systems. However, it is an expensive and not optimal way from the point of view of IT infrastructure use. Therefore, in practice, central MIS is used to support the capital management process, which is cheaper because it only reuires to create - within the existing structures of MIS - Capital Management Centre (CMC) at the bank s level and the Divisional Capital Management Centres (DCMC) at the individual business divisions level. Within the budget DCMC receive from CMC capital adeuate for planned business activities. The incorporation of capital allocation into business processes enables to consider the cost of absorbed capital at the lowest business level. This means that the individual consultant working with a portfolio of clients possesses the related information on the sie and cost of capital absorbed. In addition, preparing himself for the price negotiation, he has a comprehensive range of information about the impact of the potential transaction on the EVA for the client. The described solution doesn t reuire a very large expansion of MIS. In Figure 2 we present additional elements to be created in a typical central MIS also used to support the management by objectives [2]. One of the most labour-intensive expansion of MIS (see Figure 2) is the Capital Allocation Engine, which reuires capital allocations for particular types of risk to the transaction and customer level. The detail of the internal models used (an example of the capital allocation for credit transactions arising from credit risk is given on Figure ) is a closely guarded secret of each institution disclosed only to the supervisory authorities. The common element of models used for capital efficiency management is the consistency with the level of decentraliation of decision-making power concerning transactions exposed to the risk. In sum, the extension of a well-organied and well-performing central MIS by the capital management module should not reuire large expenditures. Moreover, the extension of MIS can be carried out by using the institution's own staff resources, as was the case in one of the banks listed on stock exchange in Warsaw (the bank does not desire to disclose its name). The use of the extended MIS system, as shown schematically in Figure 2, has allowed to involve the wider group of employees to the capital management process (at the level of budgeting, monitoring and Management by Objectives the motivation system). The implementation of this application has positively influenced the capital adeuacy ratio value. This in turn means among others the credit expansion possibility and also the advantages for shareholders, because the bank is able to spend the greater part of the profits on dividends rather than on the capital increase. 22

41 MIS Risk Managament Risk Center Risk Center 2 Risk Center 2... Risk Center n MIS Repository, input data, information Capital management module Capital allocation engine DCMC, DCMC 2, DCMC 3,..., DCMC s CMC KPI engine Capitalperformance indicators (RORAC, RARORAC, etc.) monitoring EVA customer level Capital structure monitoring External Regulatory Reporting Presentation layer and data handling processes Figure 2. Additional elements of the MIS used to support capital efficiency management taking into account the linear employees engagement 5. Conclusion In large banks, it is possible to use the more developed form of the capital efficiency management support, namely taking into consideration in the budgeting process the capital allocation to the level of Capital Management Regional Centres specially appointed within the individual business divisions. This approach allows the conscious assignment of capital within the geographical structure. Additionally, it integrates lower level managers, advisers and local stuff in their concern to increase the effectiveness of the available capital. Managers can consciously participate in the capital efficiency improvement beginning from the level of the budget and ending during its implementation, due to the information they have on both the capital efficiency (EVA customer reports) and reduction in the availability of capital. This becomes possible because MIS can efficiently provide information to be used by motivation systems. REFERENCES [] Opracowania KNF: Sektor bankowy. Miesięcne dane o rynku. (203) html, [/2/203]. [2] Odrygóźdź Z., Scesny W. (202) Practical aspects of the use of the central management information system for management by objectives in a large universal commercial bank, Information Systems in Management Vol. (),

42 [3] Scesny W., Borowiec J. (202) Functional Design of Application supporting Process of Management by Objectives in a Universal Commercial Bank, Information Systems in Management, Vol. (4), [4] Ustawa dnia 29 sierpnia 997 r. Prawo Bankowe, D. U. 997 Nr 40 po. 939, tekst jednolity (203), [.2.203]. [5] Basel III: A global regulatory framework for more resilient banks and banking systems (200), [/2/203]. [6] Regulation (EU) No 575/203 of the European Parliament and of the Council of 26 June 203 on prudential reuirements for credit institutions and investment firms and amending Regulation (EU) No 648/202 (203) [/2/203]. [7] Directive 203/36/EU of the European Parliament and of the Council of 26 June 203 on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms, amending Directive 2002/87/EC and repealing Directives 2006/48/EC and 2006/49/EC (203) [/2/203]. [8] Raport o sytuacji banków w 202 r. (203) Urąd Komisji Nadoru Finansowego, m pdf [.2.203]. [9] Matten Ch. (2000) Zarądanie kapitałem bankowym. Alokacja i pomiar wyników, Warsawa, Dom Wydawnicy ABC. [0] Pawłowic L. (20) Optymaliacja alokacji kapitału w budowaniu wartości banku dla akcjonariusy, Kwartalnik Nauk o Predsiębiorstwie, nr, [] Sprawodanie finansowe Powsechnej Kasy Oscędności Banku Polskiego Spółki Akcyjnej a rok akońcony dnia 3 grudnia 202 roku (203) [/2/203]. [2] Jednostkowe Sprawodanie Finansowe Banku Pekao S.A. a rok akońcony dnia 3 grudnia 202 (203) dania_finansowe/?r,main,tabnumber=3 [/2/203]. [3] Milic-Cerniak R., Gadomski M. (2009) Optimiation of bank s capital profitability capital allocation and EVA and RARORAC analyses at Bank Pekao SA, SAS Forum, Poland 2009, Warsaw (in Polish), [/2/203]. 24

43 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) DASHBOARD CONCEPT AND APPLICATION IN THE MEASUREMENT OF MARKETING EFFECTIVENESS GRAŻYNA GOLIK-GÓRECKA Department of Marketing, University of Lod This article attempts to bring closer idea of using a dashboard tool to measure the effectiveness of marketing activities. It presented barriers, base elements, how to create a dashboard development path. Today's dashboard solutions have the potential to collect, integrate, analyse and present reports in various contexts of the study and meet all expectations of companies and researchers. It now appears, therefore, that the companies that use the tools of marketing analytic dashboard, have the biggest chance for success on the market. Keywords: marketing, dashboard, marketing metrics, marketing effectiveness, paths of marketing metrics. Introduction To determine the most important issues of marketing research and their range, carried out by researchers and practitioners, various supporting tools may be used, such as monitoring and analysing the effects of marketing activities. These tools, using new technologies, are supporting marketing analysis in businesses. If we accept what has been stated at the Conference Controversy around marketing organied by the Department of Marketing L. Koźminski Academy, that marketing is secondary to practice, then it reuires great care to identify new directions of research and collaboration with practitioners to formulate the research problems for their needs, not forgetting about the development of marketing and the creativity of the future development paths [].

44 If we take into account only the needs of practitioners, the sphere of marketing theory will be certainly uite poor. It should be done differently, as it has happened before, for example in the United States - the famous research PIMS, and in Poland research of the companies, which was conducted together with research institutions or with the university units using experiments and interactive research. It may also be noted that these studies permeate through the cooperation of the universities and practicians, and the result is often co-developed areas of research so necessary for both parties. Other productive ventures are seminars and lectures that are conducted by both practitioners and researchers. In addition, various types of analysis and the results of joint research serves as a basis for the development of new research topics, and knowledge of young scientists can be used by the company. Therefore, it is worth even more to develop these joint actions to make them targeted and used by both sides. The tools presented below can be used for both sides, as they enrich the range of marketing analysis in practice. So does marketing analysis science supported through such measures as efficiency of operations and competition on the market this is the main thesis. 2. Need for information Stock markets around the world provide us every day with information on the current situation in corporations. Share price uotations are a source of knowledge about the current condition of the company. Unfortunately, many managers of other companies end his day without having received a response from the market. A management dashboard provides information relating to the key areas of business systems, signalling trouble before the problem is fully revealed. As an example, let s take the sales process. Most managers discover at the end of the uarter or the year that sales goals have not been met. If, however, they were using the dashboard to monitor key determinants in the sales process, they would know how effective it was from the beginning. You can then take the appropriate action. Private companies always focus on the percentage growth of the sales in plan. Unfortunately, they never give details of how they plan to achieve this growth. This often leads to failure of the intended plan, and how much different it is from what was planned, the management finds out at the end of the period [2]. This system of creating and delivering the information is outdated and very risky. 3. Dashboard how important it is for business? Console, corporate board, instrument panel or just as in the title Dashboard is an innovative instrument that supports the process of management, based 26

45 on its internal information system and database. The name of this tool and other synonyms directly relate to the specific design it is modelled on dashboards in cars as it includes set of indicators that help you organie and visualie the data necessary for the decision making process of the company. Despite the fact that this text will focus on the marketing variation of the dashboard, console, i.e. mainly filled with data regarding the effectiveness of marketing activities, it should be made clear: corporate dashboard is a multi-use tool and can also be composed of other "pads" than marketing data, so it can suit the specific operational and business strategy. Regardless of the form of organiation of marketing in the company, there is a communication problem between marketing department and the decision-making centre. It is also a problem concerning the relationship between these two centres. The decision-making centre is responsible for over-arching strategy and the direction of company development. They must have a full view of its activities and decide about the expenditure of resources (financial, human, time, etc.) which are delegated to specific actions and specific departments. The marketer, running after the realities of marketing and holistic orientation, is aware of the importance of financial capital, which is needed to identify the dynamic changes taking place in the environment of the organiation, and thus how to dynamically adapt to them. The conflict between the two centres is born at a time when the centre of decision-making puts pressure on the marketing team to provide clear evidence of effectiveness and most importantly - the effectiveness of marketing activities in relation to expenditure allocated to them [3]. The situation is complicated by the fact that decision-makers often do not have time to analyse the details (which are very critical for the strategy) of the data provided to them by marketers, or simply cannot understand such data, which often leads to short-sighted marketing decisions. This problem is solved by the premise and practice of marketing dashboard. Dashboard is an important part of the performance management. Properly constructed, it demonstrates adaptation of marketing expenses to anticipated results. It also allows the organiation to determine whether it is on the right track, leading to an increase in demand for products, improved customer retention and increased market share [4]. Dashboard is a decision support tool that facilitates strategic decision making and correcting the course of action. In order to facilitate the synthesis and interpretation, it presents internal and external key measures in an accessible and attractive way and most importantly - in real time [5]. The most important benefits of this embodiment are: Speeding up the decision making process in the company and reducing its costs; Pictorial and timely look at the situation in which the company is, resulting from the continuous monitoring; 27

46 Tracking the effects of specific marketing activities that avoid the so-called marketing myopia ; Streamlining of communication between the centre of decision-making and the marketing department; Increasing the role of marketing in the whole company and its operations. 4. Barriers and obstacles to the dashboard Some managers are concerned that by focusing on the figures presented in the form of a marketing dashboard, you can lose sight of the whole. In addition, attention shall be paid to the issue of privacy and the pressure that this techniue has on people. Most experts, however, believe that the benefits justify the risks. Pat LaPointe, consultant and analyst in the field of marketing and author of Marketing by the Dashboard Light, lists six major barriers and obstacles to undertaking of an effective measure of marketing effectiveness, and thereby standing in the way for the implementation of a well-functioning dashboard:. Applying space solutions to junk data. Even the most advanced analytical gymnastics seems to managers for nothing, if it is calculated on the basis of inaccurate or mismatched data. 2. Lack of proper proportion in the data collection process. The three guiding principles of data collecting - the speed, accuracy and relevance do not go hand in hand often. A lot of businesses choose the first one, at the expense of the other two - minimiing the cost and time of research. The effect of this approach is described in the preceding paragraph. 3. Over-reliance on IT. The people responsible for the corporation s Information Technology deal with not only the collecting and storage of data, but also the creation of profiles of the most profitable customers. De facto, they do monopolie decision-making process, making it dependent on the new technologies in a rather malicious way. Technology should be only supporting - not replacing managers or directors in the art of management. 4. Inadeuate recognition of the work of researchers and analysts. What is the exact role played by marketing researchers and analysts? It is key. What is the their role across the enterprise? Fundamental. It's their job that allows adapting company's activities for dynamically changing market environment. But what career development opportunities are waiting for researchers and analysts? None. As a rule, they receive a pittance and miserable prospects for career development. This obviously affects the uality of their work. Again - see first point. 28

47 5. Lack of investment in the skills of researchers and analysts. Lack of recognition of the uality of the work of researchers and analysts is reflected in the persistent lack of training given to them dedicated to improving their techniue, knowledge and skills. 6. What is needed is a clear leader. Delegation of power and authority to lower levels is a good thing, but it can be associated with particularities, excessive individualiation of decisions, and above all - a lack of looking at a complete picture of your business. Such an approach should have a leader representing the centre of decision-making - taking care of it all, so that everyone plays for one and the same goal. It is a leader who is able to push his own vision and lead the rest in the most appropriate direction [6]. It is worth adding to these points above a common ailment of executives, which is an inefficiency to think only in purely economic terms - out of context with hard numbers, while they are insensitive to the needs of consumers in the brand development perspective. In this case of the business philosophy, dashboard in no way proves to be a useful tool to promote governance, and will only intensify such an incorrect and out-of-dated practice. 5. Basic elements in creating of a dashboard The content f or tables may be the two most important basic elements of the dashboard, based on the observation of the market "scorecard", which are a set of metrics and provide early warning signals.. The scorecard from the point of view of the customer, which registers how the company is doing from year to year with measures of value that are important to the customer (for example, the average reputation of the product compared to the product's main competitor). Management should set targets for each indicator and take action when the achieved results appear be adversely different from the assumptions. 2. Scorecard from the perspective of stakeholder satisfaction that should monitor different groups, whose main interests are dependent on the company's performance or affects it. These are the distributors, suppliers, banks, employees, retailers and shareholders. The Management Board shall take appropriate measures, where at least one of these groups will record a significant increase in dissatisfaction [7]. The next problems is five steps to creating a marketing dashboard by L. Patterson:. Adapting marketing to business results. If you do not know what indicators will be needed for your company, everything else seems like an issue in dispute. The results that you are observing are more than just numbers. You have to understand that you want to achieve results associated with current or potential customers, to whom you have to get. You should know 29

48 how many of them should buy the product and which of your products / services should be acuired. The company can therefore use the mapping process, or other approaches that highlight the relationship between marketing programs and the results of the company. 2. Choosing your own metrics. Organiations usually have measures of the following categories: customers (acuisition, retention, value), products (innovation, price, margin), positioning (market share, brand preference), and finance (budget, return on investment). 3. The creation of chains of data between marketing activities and results of the company. Chains of data will help to show the connection between marketing activities (e.g., campaign with a call to action) marketing objectives and the company's performance, and assumptions about these three factors. 4. Acuisition of data. Measurements reuire data. The company has to know what data is already collected, and what it still missing. 5. Check and review. Please create a dashboard alpha version to test data and processes. When this version is ready, you can decide what changes (if any) are reuired and must be made. Then get a beta dashboard. Finally, you must create a pilot project, which can be later implemented. Then, it is necessary to analyse and determine how to make reporting automatic. A you can see, creating a marketing dashboard, that works well reuires time and investment, but it is worth the effort. A properly designed instrument panel contains critical diagnostic and prognostic measures and presents templates of results in the blink of an eye [8]. This is why the sphere of science is needed - analysts-researchers and students should know the ways to build alpha and beta, and reporting capabilities. This will allow to professionally carry out all marketing activities and to report and investigate. Each dashboard should be as uniue as the organiation they serve. Regardless of the physical form it takes, its aim is to create a concise and clear reporting process for defined business objectives. An effective dashboard is alive. It adapts and changes to the organiation, when the tasks crystallise and are redefined, which results in relationships between metrics. There are two main objectives of a dashboard: diagnostic insight (view) and predictive forecasting - with special emphasis on the latter. Some of the dashboard gauges are diagnostics, they observe what happens and try to discern why. The most important thing, on which we should base our presumptions, are the predic- 30

49 tive diagnostics, which use the experience to predict future performance under different assumptions regarding the circumstances and the allocation of resources. Marketing Dashboard creates a way for the unit and all people above and below it in the organiational hierarchy, to see what works, as soon as possible, creating a solid foundation for learning. Figure shows the path of development marketing dashboard. Figure. Marketing Dashboard Source: Pat LaPointe, Marketing by the Dashboard Light, Association of National Advertisers, MarketingNPV.com, USA, 2005, p. 27 Marketing Dashboard contains the following content that will derive from the card results:. Business Objectives: this is the starting point. These are the goals of the company, translated into a set of marketing objectives. All ideas and initiatives should be filtered through this prism. 2. The initiative ROI and resource allocation: an important part of the dashboard is to measure the additional cash flows generated by marketing programs and initiatives in the short term. What's more, the dashboard is a great tool to measure the efficiency of resource allocation in value of money, employment status, or both. 3. Brand and customer asset evolution: at least eual to the most important (in relation to the results of short-term) is a long-term evolution of corporate assets entrusted to marketing - usually includes brand and customer relationships. Dashboard can provide a reading of how these assets grow and how they grow. 3

50 4. Skills: refined dashboard keeps track of the skills and competence of the team of marketers in relation to the set objectives of proficiency. 5. Process: the Dashboard also allows insight into the performance of critical business processes necessary to deliver the reuired value proposition for the customer. 6. Tools: it is less a metric than an enabling factor; the successful application of the dashboard uses and constantly improves the tools to deepen the understanding and to reduce the effort, both in its production and distribution. 7. Diagnostic insight: dashboard must change the perception from what is going on to why is this happening, providing insight into where the earlier expectations were inaccurate, to help improve the process of defining the expectations and predictions for the future. 8. Predictive value: the difference between helpful and really effective dashboard, is the degree to which it utilies and provides diagnostic insight into what is likely to happen in the critical dimensions of business without intervention. 9. Efficiency and effectiveness: the ultimate goal - to increase both the effectiveness and the efficiency of marketing investments, thereby increasing the ROI and NPV for the company [9]. 6. Dashboard customiation capabilities and linkages with financial implications Pat LaPointe believes that the dashboard provides all current information necessary to conduct company business operations, such as the effectiveness of distribution channels, the evolution of the brand and the development of human resources. An effective dashboard will help to focus and organie thoughts, enhance internal communication and indicate where marketing is cost effective, and where not. La Pointe uses four "paths" used by marketers today. The path of consumer metrics - measures in what way potential customers are becoming real customers. In the sphere, it is also researched how the customer experience is reflected in the perception of value and competitive advantage of the company. The path of individual metrics - presents knowledge of marketers about the unit sales of products and services - sales by product line or region of sale, marketing cost per unit of product sold as a measure of efficiency, as well as where and how the margin is optimied in the context of the characteristics of the product line or distribution channel. The path of financial metrics - focuses on whether the marketing expenses pay off in a short period of time. ROI indicators for programs and campaigns measure the immediate impact on the net present value (NPV) of the expected profits in connection to the specific expenditures. 32

51 The path of brand metrics - monitors the long-term effects of marketing activities using measurement values that allow us to evaluate both whether the brand is healthy in terms of actual and potential customers, as well as from a financial perspective. In the LaPointe's opinion, dashboard presents lessons learned from all the paths in such a way that you can see the delicate relationship between them. A well-designed instrument panel should have tabs which will group the various "family" of measures arranged by product, brand, experience, distribution channels and macro-factors. Figure 2. The many Roles of the integrated Marketing Organiation Source: Pat LaPointe, Marketing by the Dashboard Light, Association of National Advertisers, MarketingNPV.com, USA, 2005, p. 26 The meaning of fully integrated marketing is collaboration of all the aspects of marketing in all the departments of the company, to achieve better results for the business as a whole. For this goal to be met, marketing has to speak the same language as the rest of the departments the language of metrics and numbers. A newly created dashboard can be a good start of integrating marketing into the context of the company and redefining its role. 7. Practical sphere Dashboards are made by such companies such Domo, Dundas, SalesManago and Outbox. Polish companies (the last two) only create prototypes or focus on virtual business (e-marketing, website traffic, SEO). It does not give a complete 33

52 picture of the leadership, which we wrote about earlier, but for companies selling online can be a remarkable tool. SALESmanago is one of Europe's first - class marketing automation systems. SALESmanago, the Polish system, is the fastest growing platform of this type in Poland. More than 400 companies in Poland and Central Europe use the system SALESmanago Marketing Automation. Companies that carry Marketing reach more than 400% increase in the effectiveness of s and convert sales. The system is used by over 400 companies, among others. Rainbow Tours, SITA, Eden Springs, Audioteka, Cinema, Jean Louis David, Answear.com and a large group of small and medium-sied enterprises. SALESmanago Marketing Automation offers the most specialied and innovative suite of tools in the field of marketing automation. The system includes a tool for the identification and monitoring of people on the web and social media, CRM, E- Mail Marketing, form generator and landing pages and dynamic content on the website. Outbox - the polish system is used by English J Bross, Spring source Drods, Oracle, Hibernate. He specialies in IT consulting, Enterprise Product Management and implementation of CRM solutions for example in Telco & Media, financial services and education services. Dundas the system is used by - Astra Zenecca, Black Berry, Coca-Cola, University of Utah, Teck and in Poland - exclusive distribution partner Passus Sp. o. o. - system is used by large enterprises from List of the 500 largest companies. Figure 3. Executive Dashboard Source: Materials provided by Dundas Company with a permit to publish 34

53 There are in dashboard marketing four paths and three platforms, there is a lot of ways of reporting. In addition, it is possible to diagnose - to plan, create prospects and monitor performance. Scheduling should be based on market research and forecasting, so you can assess how the market is in a perspective of time, how can be structured and what the value of customers will be like. Based on the historical data can be answered not only in short term, but what will happen in a year. Dashboard clearly presents charts and indicators, which are grouped in various tabs. Each of these contains generally 4-5 metrics. Often these are charts, rates, maps (in terms of sales regions), and indicators like tachometers in the car. Same tab (usually four) is based on the needs of the enterprise and concern issues such as sales, support, and products. Below is an example of a dashboard. REFERENCES [] Plenary discussion on the Conference Kontrowersje wokół marketingu. Kreowanie wartości a wartość marketingu, ALK, Warsawa, ora C. Misiewic, Sprawodanie Konferencji. Handel Wewnętrny, IBRK, KiK, Nr 3, 203, pp [2] M. Celli, What Every CEO Dashboard Should Be Tracking, MarketingProfs, [3] Ph. Kotler, K.L. Keller, Marketing, Pearson Education, wyd. 4, Ponań202, p. 25. [4] L. Patterson, How to Create an Actionable Marketing Dashboard (and Why You Need One), "MarketingProfs", [5] Ph. Kotler, K.L. Keller, Marketing, Pearson Education, wyd.4, Ponań, 202, p. 28. [6] P. LaPointe, Six Primary Obstacles to Marketing Measurement, MarketingProfs, , [7] Ph. Kotler, K.L. Keller, Marketing, Pearson Education, wyd.4, Ponań, 202, p. 28. [8] L. Patterson, How to Create an Actionable Marketing Dashboard.... [9] Ph. Kotler, K.L. Keller, Marketing, Pearson Education, wyd.4, Ponań, pp , a: P. LaPointe, Marketing By the Dashboard Light, Association of National Advertisers, 2005, 35

54 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) PROPOSAL OF A NEW METHOD FOR E-BANKING AUTHENTICATION BASED ON FINGERPRINT READER ARTUR HŁOBAŻ, KUBA OWSIŃSKI Department of Physics and Applied Informatics, University of Lod The paper presents a novel method/mechanism for electronic banking transactions authentication based on biometrics, using fingerprint reader. The method allows to identify the client not only by something that he has or knows, but who he is. Therefore, the proposed solution enhances security of transactions authentication by an additional factor of client identification - this kind of solution hasn't been currently used by banks in Poland so far. The article also presents that the application of an additional level of security is not associated with an increase of costs for banks but allows them to reduce them, especially for long-term relationships with customers. Keywords: Electronic banking, Authoriation codes, Authoriation, Biometric data, Fingerprint scanner. Introduction Banking, like many aspects of our lives, moved to the Internet. Clients are increasingly willing to use this method of payment and control of their finances because of the convenience and time savings []. However, with the popularity of the Internet and electronic banking has increased the number of attacks and extortions by online criminals [6,7,8,9]. Hacking attacks on bank servers are just one of the methods. Statistics show that much more effective attacks are directly in the client. It mostly often happens through e.g. substituting a fake login webpage of the bank and extortion of authoriation data. Banks increase security by encrypting connections, masking passwords or detection of failed logins. However, the most effective

55 to approve operations with a high degree of risk are authoriation codes, such as transfers to external clients or changing personal information of the client in the bank system [2]. The purpose of this article is to present a new method of banking operations authentication in the telecommunications banking services area. Due to the diverse needs of retail and corporate customers in the article will be taken into account only the services offered to individual customers. 2. Device selection For the tests, the Transcend JetFlash 220 4GB device was selected, which already includes software to read the fingerprint pattern and validate it. Manufacturer software protects access to the disk using 256 bit key (AES) [5]. During the first start up of the device the client is asked to scan a fingerprint and after that while every next starting the device will ask for moving the finger through the scanner and will verify fingerprint. Only after positive comparison, the client will have access to the device on which the generating codes application is stored [3, 4]. The authentication system consists of a client application running on the client machine, which is activated from the JetFlash 220 device, and application on the server running on the bank side. Client application does not perform data validation but verify only their format and their length. After that the data are sent to the server and client receives back the answer. 3. Client application The task of the client application is only to obtain relevant data from the client and send them to the server (Fig. ). The data, which will be sent, are: ID uniue device serial number. This number will be read each time from the device's memory. It is essential to the production stage, that this number has to been stored in device memory with no possibility of modifications. It will be read from the device and assigned to a particular user at the time of delivering the device. CIS the same number as the ID number in electronic banking. The customer will have to enter it every time because there will not be write access to the device after the production process. Entering CIS number will also provide additional protection. PIN access code which provides additional protection. PIN will be determined by phone call at the time of authentication method activation and will not be stored in memory. 37

56 CLIENT IP IP address of the computer from which the reuest was send. DATA OF OPERATION data set by the bank being characteristic for specific transaction. This could be the last four digits of the account numbers of customers, which are presently used by banks. Data is encrypted and sent to the server and the client program waits for a response. Figure. Diagram of the client application - a single code generation 38

57 This method is useful for the approval of transactions on the website because the customer who is going to make the transaction in such a way, must have access to a computer with Internet. In addition, it is possible to generate not just one code, but also a list of codes. 4. Server application The server waits for a response, ensuring uninterrupted operation and support for multiple clients simultaneously. The application receiving a client reuest first check whether the customer actually performs an operation that reuires authentication. The purpose of this is to protect the server from the load due to the generation of unnecessary codes. Next is being verified whether the customer has activated the biometric tool and the CIS or the ID of the device is not blocked. A further step is to verify that the client uses its own device and introduces the correct PIN. The last step is to check the transactions data and generate an authoriation code (Fig. 2a and 2b). The similar schema of action is in the case of receiving a reuest of code list generation. In addition to data validation is to verify the expiration date and the number of codes that the client wants to receive. This is the stage of generation, which can be freely modified by the bank. 5. Method of device activation Device delivery to the customer can be done by sending it from the bank's head office directly to a particular client or to individual bank branch where the client can get it in person. The second method is preferred for two reasons. Firstly is cost, because of the small dimensions of the device it will be cheaper for bank to deliver package containing several devices to a single address (bank branch) than the delivery of each device to a separate addresses of clients. Secondly there is the need to wait by the client for the delivery while the bank's branches can create supplies and give the device, for example, in the moment of opening an account by the client. Regardless of how the euipment will be given to the client, it should be preassigned to him. The proposed method is coding ID of device in the bar code. The bank clerk before issuing or delivery the device, scan the bar code and into this way assigns it to the client. The label with the bar code should be peeled off before giving the device to the client for safety reasons. 39

58 Figure 2a. Diagram of the server application if it receives a reuest for code generation 40

59 Figure 2b. Diagram of the server application if it receives a reuest for code generation 4

60 On the telephone number, specified by the customer in the bank facility, contact call center employee who verifies the client (for example, by logging in to the telephone service) and sets with him the PIN to a biometric tool. The client enters the PIN on the phone, which is stored in a database. This method of PIN generation is nowadays being used by banks to determine PIN to hardware token, credit card or telephone services. During the first start of the device client folds the fingerprint pattern, runs the application and generates the first code, which enters on the bank website in the section of the authentication settings. If the generated code is correct, biometric device is activated and becomes the only way to authorie the operation of the client. 6. Comparative evaluation of the security Nowadays offered methods have different levels of security. They can be based on what the customer has (e.g. the card with one-time codes, cell phone with number on which SMS with code comes) or/and knows (e.g. PIN code). The third level of security is based on a hidden data. In the case of a hardware token client must have a device, knows PIN and the generation of the code is done on the client and server side on the base of serial number stored on both sides, however not known for the user - it minimies the risk of extorting this data from the client [2]. The proposed solution introduces a fourth level of security, which until now has not been used to authenticate banking transactions. In summary, the security of authoriation code generation is based on four characteristics:. something that the client has - biometric device, 2. something that the client knows PIN, 3. something that is assigned - the uniue ID of the device which is unknown for the client, 4. something that confirms the identity of the client fingerprints. Clearly it can be concluded that the proposed method offers the highest levels of security and the information on which the authoriation codes are generated. 7. Comparison of the costs of methods Because it is not possible to estimate the cost of a single code generation for every method, the best indicator of the cost will be incurred in generating a number of codes in a given period of time. The following simple analysis of the expenses of each authoriation method for one user will be made in the period of one (Fig. 3), twelve (Fig. 4) and thirty six (Fig. 5) months. A generalied value of 20 42

61 operations per month was assumed for each of them. It is not excessive, especially since more and more banks will reuire an authoriation code not only to approve financial transactions, but also for the login to the website, for example to see the account balance. On the basis of analysis of the charges in selected Polish banks [0,,2,3,4] assumed average prices for the most popular authoriation methods nowadays: - Card of codes - issue the card of codes for 50 operations is an expense of 5 ł [2]. - SMS codes it was assumed that the cost of one SMS is 0.20 ł (0.20 ł x 20 operations = 4 ł) [, 3]. - Hardware token - generates a fee for the issuance of 48ł and ł for monthly usage [5]. - Biometric token - the purchase price of a single device of Transcend JetFlash 220 4GB for retail client is 5,30 ł ( January 203). It is important, that the costs can be further reduced by a significant reduction in memory capacity and mass production. It was assumed that the customer will not be charged monthly fees by the bank for the use of biometric token. 60,00 ł 49,00 ł 5,30 ł 50,00 ł 40,00 ł 30,00 ł 20,00 ł 0,00 ł 5,00 ł 4,00 ł Card of codes SMS codes Hardware token Biometric token 0,00 ł Figure 3. Costs of month using of chosen authoriation methods for one customer 43

62 60,00 ł 60,00 ł 48,00 ł 5,30 ł 50,00 ł 40,00 ł 30,00 ł 20,00 ł 25,00 ł Card of codes SMS codes Hardware token Biometric token 0,00 ł 0,00 ł Figure 4. Costs of 2 monthly using of chosen authoriation methods for one customer 60,00 ł 44,00 ł 40,00 ł 20,00 ł 00,00 ł 75,00 ł 84,00 ł Card of codes SMS codes 80,00 ł 60,00 ł 5,30 ł Harware token Biometric token 40,00 ł 20,00 ł 0,00 ł Figure 5. Costs of 36 monthly using of chosen authoriation methods for one client 8. Conclusions This paper proposes a new method for banking authentication that uses a fingerprint reader. It is important that the new solution provides a higher level of security and convenience of service, while reducing costs in the long term use. The implementation of biometric token by any of the banks will add not only additional authentication method to choose from, but the solution cheaper, safer and higher level of prestige and functionality. 44

63 REFERENCES [] Matusyk A., Matusyk P. (2008) Instrumenty bankowości elektronicnej, CeDeWu Centrum Doradtwa i Wydawnictw. [2] Anderson R. (2006) Inżynieria Zabepieceń, Wydawnictwa Naukowo Technicne. [3] Bolle Ruud M., Connel Jonathan H., Pankanti Sarath, Ratha Nalini K., Senior Andrew W. (200) Biometria, Wydawnictwa Naukowo Technicne. [4] Ślot K. (2008) Wybrane Zagadnienia Biometrii, Wydawnictwa Komunikacji i Łącności WKŁ. [5] Menees A., Oorschot P., Vanstone S. (2005) Kryptografia Stosowana, Wydawnictwa Naukowo Technicne. [6] Portal Niebepiecnik.pl (stan na dień r.) Phishing na klientów ING, [7] Portal Niebepiecnik.pl (stan na dień r.) Phishing na klientów BZ WBK, [8] Portal Niebepiecnik.pl (stan na dień r.) Atak Zeusa na klientów bankowości internetowej, [9] [9] Portal Niebepiecnik.pl (stan na dień r) Kradież danych tokenów RSA, [0] [0] PKO BP, Tabela Opłat i Prowiji [] PEKAO S.A., Tabela Opłat i Prowiji, CHMENT,PL,,20,0/Nowa_linia_rachunkow_oscednosciowo-roliceniowych.pdf [2] MBank, Tabela Opłat i Prowiji, [3] Inteligo, Tabela Opłat i Prowiji, [4] ING Bank Śląski, Tabela Opłat i Prowiji, [5] PEKAO S.A., Oferowane metody autoryacji, /#tab2 [6] Ostrowski E. (stan na dień r.) Metody i Algorytmy Stucnej Inteligencji, 45

64 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) CHATBOTS FOR CUSTOMER SERVICE ON HOTELS WEBSITES MIROSŁAWA LASEK a), SZYMON JESSA b) a) Department of Information Systems and Economic Analysis, Faculty of Economic Sciences, University of Warsaw b) Denise Systems sp. o.o. In this article we present an analysis of implementations of a chatbot - a program which simulates an intelligent conversation with webpage visitors, dedicated to hotels and guesthouses (hotel chatbot, in short: HC). We obtained uniue data from five various webpages exhibiting various configurations, containing a total of 743 user statements in 465 conversations. HC informative function was confirmed for more than 56% of the conversations. Moreover, 63% of users prefer to interact with HC if it suggests at least one clickable option to choose as an alternative to typing. The results indicate that the implementation of speech synthesis increases the percentage of users who decide to start a conversation with the HC and it may have a positive impact on the percentage of users that book rooms online. Keywords: Reservation Systems, Hotel Chatbots, Hotels' Websites, Touristic Services, Human - Computer Interaction, Economic Value of Chatbots. Introduction The term chatbot relates to a computer application with which users can conduct a dialogue in natural language as if it was conducted with another person []. The most widely described prototype program, which falls into this category, is Elia, which provided the responses of a nondirectional psychotherapist in an initial psychiatric interview [29, 30]. Elia's success is often explained by the fact that users uneuivocally anthropomorphied and somehow set up a relationship with the program, often highly emotional [2]. The term "Elia effect" describes

65 this human tendency to assign chatbots the attributes of human intelligence [3]. Nevertheless, the truth is that Elia and most chatbots use various tricks to simulate intelligent behavior and to make an impression of a speaking human being. This goal distinguishes it from programs that also use natural language interfaces, but have more advanced algorithms used for text analysis and reasoning, while the aspect of human-like naturalness of communication is marginalied [5]. Tricks used in chatbots are based on algorithms that use textual pattern matching rules and dialogue management rules [0]. Perhaps the most surprising fact is that these simple methods can induce powerful delusional thinking in uite normal people [29]. It may be due to the fact that in the history of mankind speaking concerned only the human species, although the idea of speaking with other beings has accompanied people since the dawn of history and finds its expression in literature and religions [2]. Despite this, we unknowingly anthropomorphie computers and treat them as social actors, though to a limited extent [2]. Chatbots are looking for their place in e-business [6]. Some of them are used on websites as shop assistants [8], helping to choose products from the company's offer and causing company s sales growth. In some cases, chatbot authors report a 8% increase in purchases, and others indicate that 5% of people who chatted with sales assistants made purchases [7]. Chatbots are designed to assist users in navigation on the site, to limit the amount of clicks and to shorten the time to reach the desired information or product. Chatbots also give other benefits, in terms of building social relationships with customers, increasing customer confidence in the company and strengthening a customer's emotional bond with the company [8, 7, 23]. More and more industries relocate their marketing, sales and maintenance services to the Internet. One of them is the hotel industry, where it is estimated that 75% of purchases are made through on-line booking systems [3]. A new tool that supports the hotel industry is a chatbot (HC) offered by Denise Systems. It aims to meet the rigorous demands of the mass market: HC implementation in the new hotel must be made as simple and automatic as possible. In result, the implementation process is reduced to the following: the hotel owner gives details on the form, chooses the chatbot look, specifies its location on the screen and pastes it on the website. HC has been programmed in order to perform specific business goals. It is not a human-oriented simulation and it can instruct the user about the scope of its functions. HC does not aim to win the Loebner [9], it is solely a targeted marketing tool. Knowledge base was limited to the hotel industry and tourism topics, therefore it does not contain elements of so-called "small talk". HC gives more information about the hotel than a user can find on a website, it helps users to navigate the site, promotes the building and surroundings of the hotel and gathers valuable marketing data from users. In this article we present an analysis of conversations between users and HC. We carried out our analyses in such a way that the results can be compared with Max male chatbot at Hein Nixdorf Museums Forum in Paderborn (Germany), 47

66 and partly with the results concerning other chatbots: Sgt. Blackwell (male chatbot at Cooper Hewitt Museum in New York), Talk-Bot (robot chatbot available online), Bill (male chatbot available online) and Kathy (female chatbot available online). 2. Database of conversations Data were collected from 4 websites of guesthouses (A, B) and hotels (D, E) that implemented their own HC, and from the HC demonstrational website of a fictional hotel (website C). All implementations are almost identical: the knowledge base is in Polish language and differs in ca. 00 responses about the hotel, which represent approximately 7% of all responses in the HC knowledge base. Location of the HC on the webpage and its appearance may vary from one implementation to another. All conversations were recorded with date, time, user input and HC answers. Logs do not contain information whether the user input has been written or selected by clicking an option. If user input corresponded to the expression accessible in option, it was assumed that the option was clicked. Inputs were tagged with categories by the most recent HC version available when writing this article. However even the latest HC was unable to find any suitable category for 22 (0%) inputs (for comparison: Max was unable to recognie 25% of inputs [22]). Among these 22 inputs, 88 contained uniue content, of which ca. 300 (.7%) were assumed random keystrokes (in Max: 3.% [8]). Among others, some inputs were in English and Russian, there were also system commands, URLs, typos and spelling errors. There was also a very small group of expressions which indicated deficiencies in the knowledge base, for instance: several uestions about maids, virtual tour, room sies, vouchers, possible discount negotiations, availability in a given date, job/internship offers, insults, vulgar behavior, sexual comments and orders given to the system (like raise your hand etc.). In total, the database analysed contains 743 user inputs, of which 226 were written, and 5287 were selected by clicking an option. Of all the user responses only 5957 expressions contained uniue content: 4955 expressions were used only once and 002 were repeated (2458 times). If the system is able to interpret the 7% (002) of inputs repeated during conversations, it will thus be able to interpret 70% of all users responses. This indicates that a chatbot designed for a particular field with limited topics and uestions range doesn t need to have a large knowledge base, and in most situations it will work properly. For comparison, the uniue content chatbot responses were 34, of which 35 were used once and 990 were repeated. The system uses a limited number of responses, therefore it was tested for repetitions that occurred during one conversation. We observed that: 2633 (63%) conversations contained repeated expressions, and 62 (39%) conversations had more than one repetition of an expression. Repeated statements were concerning: prices 48

67 presentation, standard rooms descriptions, descriptions of single rooms and double rooms, room euipment and responses in situations when the HC was unable to determine the category for user input. Average user input length was characters, and for HC answer 80 characters (for comparison: on Polish IRC network, most responses consist of 0 characters). 8% of all registered responses were threecharacter expressions, mainly yes 778 occurrences, and no 777 occurrences. Similar results concern the IRC network, where the words no, it and yes are the most popular as a one-word expressions [4]. The fact that chatbot responses are much longer in comparison to the user responses, was observed as well by Max authors [8]. 3. Conversation length and duration Conversations were conducted via a web browser, and users were identified by cookies. Conversation duration is counted from the first user s expression to the last statement of a chatbot. The pause between the user s statements cannot be longer than 5 minutes if the same user re-enters anything after 5 minutes, a new conversation start is assumed. The average conversation length: 4.2 user inputs, 8 seconds. On average, the longest conversations observed on website C 7.7 user inputs per conversation and about 6 minutes (368 seconds) average duration. It may result from the fact that HC on this website is a demonstrational implementation, which serves for testing and presentation of the HC itself. Note, that chatbot on website D has few conversations, therefore the averages may significantly change in the future (this also might impact the standard deviation of number of user inputs, greater than in other implementations). It is interesting that the website A has mainly very short conversations. This may be related to the construction of this webpage, because HC is placed on top, and almost all of its pages reuires vertical scrolling, which causes the HC disappears from the screen. It may be the main cause of short talks on website A. In comparison with other chatbots: () Sgt. Blackwell 4 inputs/conversation [24], (2) Max inputs/conversation, duration: 3 minutes [6, 20], (3) Extempo Sales Assistant 5 inputs/conversation, duration: 2 minutes (in 90% of conversations) [2], (4) Talk- Bot inputs/conversation [5], (5) Bill inputs/conversation [5], (6) Kathy 3.63 inputs/conversation [5]. Conversations with HC are relatively short in comparison with other chatbots. One reason for the discrepancy may be that the HC is only an addition to the hotel website, and other compared chatbots stand mostly in the foreground. Looking at the website C, where the HC appears in foreground, it may be assumed that the role of the chatbot in foreground or in background, affects the length of the conversation. Another important reason may be that other chatbots often have a very extensive knowledge base for small talks which is not present in HC. Despite this, we registered 44 (%) conversations 49

68 which were longer than 0.5 hour. Sheryl Brahnam has examined that conversations conducted with the female chatbot Kathy are longer than those conducted with male chatbot Bill and Talk-Bot robot. On the website D we observed the opposite conversations are longer than those on A, B and E webpages, which use the female chatbot. Only the implementation of webpage C has longer conversations, but its chatbot can be female or male. Max authors think, that the online conversations will be longer than those conducted with Max, because interlocutors were probably sitting in front of their computers having a private chat. ( ) In the museum, the users are standing in front of Max and his answers are spoken out loud [22]. 4. Typing and pointing HC gives the user the choice of entering a reply, or selecting an answer by clicking it from displayed options. In this way, the HC is getting closer to the realiation of the idea that the best way to facilitate Human Computer Interaction is by allowing users <<to express their interest, wishes, or ueries directly and naturally, by speaking, typing, and pointing>> []. From the total 6607 clickable responses displayed, users clicked Additionally, it was verified that 267 (63%) talks were initiated by clicking an option displayed in the HC welcome message. This indicates that users are much more likely to interact with a chatbot when they can click a prepared response. Users who chose to click, are more likely to continue the conversation, if the next response also gives them this option of communication. Users who chose to click, more often abandoned their conversation than users who chose to write. Users who chose to write, often abandoned conversations, if a chatbot displayed them options to click. Conclusions are as follows: () specifying the option to click increases the chance that users will interact with the chatbot, probably because clicking reuires usage of the mouse and clicks, and therefore is easier to do than typing in the text; (2) chatbot should adapt its interface to the user, and if the user prefers to click, then chatbot should guarantee him another option, and if the user prefers to write, then chatbot should restrict displayed options; (3) ultimately, the options should be attractive enough to draw the user into the conversation and convince the potential customer to write, which is even more engaging than clicking. There are two categories of statements, which more often resulted in clicking: () uestions such as where did the user find the information about the hotel: internet, media, friends; (2) presentation of pricing and room choice reuest: single room, double room etc. A huge preference to click the option in the short statements containing maximum three options was observed. This may indicate that users prefer options over which they do not have to think long and do not need to read or analye the context of the whole expression. 50

69 5. One-input conversations analysis We registered many conversations, which were very short: 508 (36%) conversations consist of only one user input and 2506 (60%) conversations had no more than two user inputs. The best results are on website C, where only 34% of conversations last no more than two user inputs (for comparison: Max 47% - no more than 2 user inputs [22]). Among all implementations, the least successful is website A, where nearly half of the conversations consist of only one user s input. The reason for this may be the previously mentioned need for vertical scrolling to view the content of the page prices, booking form, or gallery, and it hides the HC. It is also interesting that as many as 24% of users engage in conversation with the implementation of the test page (website C) lasting at least 0 user inputs. In the case where the first input results from a click, and the HC response does not give the subseuent option to click, 38% talks were continued and 62% were ended. It means, that 789 users did not decide to go on writing, when the HC didn t display any further clickable options, and that s the case in 52% of all one-input conversations. In total there were 255 responses (used 508 times!), however subseuent responses contain much less occurrences. Six most common responses contain over 50% of the endings. These are also the most common expressions occurring as the first HC response. Responses without options causes freuent resignations from the conversation. The option scheme is proposed to continue discussions in a certain direction. If users do not have their own idea, they will often follow the chatbot direction [27]. And if a chatbot does not propose an option, the conversation ends just before the user writes anything. 6. Conversation content analysis One of the tasks of the HC is to provide answers to uestions concerning the hotel s offer. We examined the extent to which all users enjoy the chatbot as a tool for obtaining information about the hotel. To this purpose, we have counted users uestions of a given category and conversations in which at least one user utterance was a uestion about the hotel/products/attractions. While analying a single response, we ignored the context of conversation (for instance the user answered yes or clicked an option to access information about the hotel without actually writing the uestion). The results should be interpreted that at least 56% of the conversations contained users uestions about the hotel and/or its offer. This kind of ueries constituted 7.2%, and for comparison Max received 2.3% of ueries concerning the museum. In addition, we tested the interest of users towards the chatbot. It turns out that 2% of the conversations include uestions of users concerning the chatbot. The number of such ueries is a total of 7.2% (260). For comparison, Max received 4.6% of such ueries [8]. This means that users 5

70 are also interested in knowing the HC, though to a lesser extent than Max. This is probably due to the fact that HC preferred and suggested topics related to the hotel and refused to talk about unrelated topics. We deliberately distinguished who is it category from what is it, because the first uery assumes a person, and the other a thing. Majority of users perceived chatbot as a person, which may be related to Elia effect [3]. This effect of anthropomorphism can have negative conseuences and generate strong negative reactions from the part of the user [2]. Further part of this publication explores these issues deeply. 7. Human-Chatbot relation The creators of Max noticed that people are likely to use human-like communication strategies of beginning/ending conversations like greeting and farewell [8]. We examined whether users greet or say farewell to the HC. Both Max and HC greet the user, therefore, according to the authors of the Max, greetings can be triggered by the greeting of the agent. Results of other chatbots: () Max - greetings 57.6% conversations [8], (2) Max - farewell 29.8% conversations [8], (3) Sgt. Blackwell - greetings 5.6% [24], (4) Sgt. Blackwell - closing.8% [24]. These results show that HC has more greetings than Sgt. Blackwell and similar amount of farewells. Max has the biggest number of greetings and farewells. Among all implementations, the biggest number of greetings - 3% of conversations with the greetings, had implementations C and D. In addition, we analyed the users statements concerning their relation to HC - whether they liked it (e.g.: "I like you", "You're cool," "You are very pretty!") or evaluated negatively its characteristics (e.g.: "but you're ugly"). The evaluations ignored behavior regarding sex and vulgar and offensive behaviors that are examined later in this publication. However, statements evaluating the system as stupid or bad (e.g.: "You are stupid") were taken into account. Altogether, statements including positive and negative evaluation constitute about %, which is two times less than reported by the Max authors (i.e. 2%) [8]. However, for Max the ratio of statements of positive evaluation to negative evaluation is.6: (5:32) [8], the HC's ratio is.3: (00:79). In addition, the amount of conversations has been measured, which contain an evaluation. Most conversations with assessments occurred in implementations C and D, which have also the longest conversations. This indicates that the assessment takes place in subseuent stages of the conversation and it takes time for user to begin to express his assessment. The most positive evaluation concerned the appearance and in second place the liking. On the other hand, it was the intelligence of a chatbot that was assessed negatively the most often (e.g.: "you are stupid") and on second place - its appearance. 52

71 8. Abuse and sex talks In a society, the aim of abuse is to cause some form of suffering to the recipient. Although chatbots do not feel, they constitute the aim of such attacks [4, 9]. We have observed that people are much easier to behave in this way in relation to the robot/computer/chatbots than to other people. Especially when user recognies that he is talking to chatbot, and not the real man or woman then clearly changes the way of behaving towards chatbot [25]. It is possible that this is a mere curiosity of users and their desire to test the chatbot. Often this is also the behavior of users represented by the minors who would otherwise never have dared to expose adults at anything so offensive [22]. This could be also explained by the human tendency to dominance and being rude which reflects a relationship in which users want to be as a higher race, where the user (man) is the master, and the chatbot (computer/robot), the slave []. HC politely guides users on topics related to the offer of the hotel, and if necessary, it explains its limits and explains in what it can be helpful. The goal is to raise users awareness of the possibilities of using HC. By engaging users in conversation, HC disperse their potential aggression and improve the uality of the conversation with the chatbot. This seems to be a good approach to be used in chatbots [6]. Some researchers complain that the negative reactions of users are ignored in the literature concerning chatbots [0]. To increase the research value of this publication, we carried out the analysis in this direction. Assuming that this category of statements contains vulgar, indecent and insulting vocabulary addressed to the chatbot (e.g. I do not know this, and you still tell me that you fucking know it, what the fuck are you telling me?, fuck off, etc.), we measured the freuency of appearing such a vocabulary in conversations with the HC. It turned out that the numbers are relatively low i.e. only 2.3% of all statements gathered in 4% of all conversations contained vulgar vocabulary. Most verbal abuses were observed in implementation D 5.8%, then C 3%. The cause may be similar as in the case of evaluations of the system conversations lasting longer encourage users to go beyond the main functions of the chatbot. Report on the chatbot Sgt. Blackwell confirms it, because it also has short conversations (4 inputs/conversation) and only 3% of the inputs with abuses. Another reason may be the form of video chatbot the only male character is implemented on the webpage D. For comparison, in other chatbots the percentage of such statements of this nature is higher and ranges from 3% to almost 7%: () Sgt. Blackwell 3% inputs [24], (2) Max 5.4% inputs [8], (3) Kathy 6.72% inputs [5], (4) Bill 6.50% inputs [5], (5) Talk-Bot 6.0% inputs [5]. Chatbots certainly must be designed to deal with abuse caused by users. Their success may depend greatly on their ability to handle users verbal abuse [2]. Chatbots also offer the possibility of trying various scenarios of conversations and behaviors, including those relating to sexual behavior. The most important thing is that all these attempts are entirely 53

72 safe, do not hurt anyone and are often the only opportunity to experience a variety of roles for the user. This causes the appearance of a series of conversations concerning sexual expressions [0]. Statements containing erotic and pornographic expressions, included those on sexual proposals addressed to the chatbot (e.g.: shall we shag now?, Let s meet for sex ) or other vocabulary related to sex (e.g.: orgasm, show spa and tits, are you gay? ). Percentage of statements of this nature ranges from 0.5% in Max [8] (which is publicly available at the museum and you have to speak in public to communicate with Max) to more than 8% for female online chatbot Kathy (the others: Bill 9.76% inputs, Talk-Bot 2.49% of inputs [5]). Sexual expressions addressed to HC measured only.8%. This may indicate that the context of a real company, a hotel, which perhaps will be visited by a user, has a greater influence on the polite behavior among users. Another explanation may be the chatbot construction assuming always returning to the main topic concerning hotel offer and not allowing for small talks, as it is a case for other chatbots. In addition, the results do not confirm unambiguously the assumption that more conversations containing sexual expressions concerned chatbots with female embodiment [5]. It is true that most conversations about sex were observed on webpage C 7% and this is the female chatbot, nevertheless the male chatbot on webpage D experienced almost the same number of statements concerning sexual expressions. To sum up, we would like to present the opinion of Peter Wallis, who writes: the problem is not to make a machine that is accepted, but to make it behave itself once it is accepted as an actor in the appropriate social context. ( ) Abuse is the fore runner to actual harmful action and as such leaves space for individuals to change their anti-social behaviour. Whereas humans and puppies are hardwired to know what these second order behaviours mean, Aibo s and chatbots need to be told. This is the challenge, I believe, that stands between us and the creation of effective human-machine conversation. [28]. 9. Economic value It is very difficult to measure the economic value of chatbots. One of the methods of measuring their value in business is an analysis of changes in the sales in online stores. For hotels, one can measure the change in the percentage of webpage visitors that booked rooms online. As the tourist market is a seasonal market, the data should be compared to the year in which the chatbot was not present on the website. Unfortunately, obtaining such data is difficult and takes time. Due to the lack of access to necessary data of all hotels, we were able to examine the economic importance of only one HC. We checked the conversion rate, which is the amount of webpage visitors compared to customers that made bookings online through special form. Additionally, we also examined what was the effect of implementation of the speech synthesis for the chatbot. We compared the periods 54

73 before the implementation of speech synthesis and after its implementation. The results show that in 20, when the HC was installed on the webpage, conversion was substantially higher than a year earlier. This applies to both the chatbot without speech synthesier (in the period March-April) and with speech synthesier (in the period May-August). The measurements were made on this website, which had a relatively small number of visitors, therefore the results should be treated with reserve. This kind of experiment should be conducted on a wider scale to measure accurately the effects of HC on the amount of reservations. Nevertheless, it seems that using the HC may cause increase in revenues, and such a hypothesis is definitely worth further investigation. Examining the impact of the speech synthesier in the period from May to August, it turned out that after the implementation of speech synthesis, conversations parameters didn t change the number a user statements in one conversation remained unchanged (the difference was at 0.03 level of expression). However, the percentage of visitors that started a conversation with a chatbot also increased by about 5%. And this indicates that the speech synthesier encourages users to start a conversation, but has no significant impact on the conversation parameters. If, indeed, speech synthesier has only such an impact, it could be applied only to the first welcome message of the chatbot. Moving on the synthesier could be easily replaced by any recording, the aim of which would be to attract attention and inspire visitors to start a conversation with HC. These aspects of the implementation of the HC should be explored in future. As a tool for providing information to users, the HC supports the hotel functioning in the field of customer service [6]. On the other hand, analysis of conversations with HC allows owners to know the needs of customers and their opinions on the object. During its functioning, the HC allowed hotel owners to detect several gaps on their webpages, e.g.: the distance to the beach and airport, availability of towels; it also pointed out that users were mostly interested in double rooms and collected a number of opinions on the webpage and hotel. Some of the gathered information is difficult to obtain in any other way. One of the uniue HC features is that it can talk with all visitors simultaneously, and even if they do not execute the purchase, it will store valuable information why they didn t choose the offer. From a business perspective, this information is priceless. 0. Summary In science it often happens that the findings are pursuing other findings, and the whole world waits long for their implementation in everyday life. Thanks to our analysis of HC we have the opportunity to provide an additional amount of data from the implementation of chatbots in the "real world" for several deployments simultaneously. The results show that the HC fills its function of an information tool. In addition, increasing the percentage of booking online in one of the hotels 55

74 indicates the impact on sales growth. Thus, the HC confirms its economic value. The results are partly consistent with the results of other researchers in terms of user interest in the chatbot, the occurrence of abuse and sex expressions, which confirms also treatment of chatbots as social actors. The resulting differences are a good starting point for further research, and new data collection, such as the responses selected from displayed options. The measurement of differences in the implementation of speech synthesis should be also verified and compared by other researchers. It is clear that chatbots do not understand what users say in that sense in which people do it, and we can distinguish a number of limitations which concern chatbots compared with human intelligence [5]. Nevertheless, the Chinese proverb says that even the longest journey (in uest of true artificial intelligence) begins with the very first step. We already know that to achieve our goal we will need many steps. Chatbots are one of them, though they certainly do not represent the whole progress in the field of artificial intelligence and cannot be a measure of the level of that progress [26]. The final verification of the discovery is an experience that confirms or refutes it and reveals new data for creating new theories. Introduction of chatbots to everyday life, and especially to the business, may be perceived as such a verification, and provides us with new data by which we learn about the social role of computers and our own perception of what is a prelude to the true artificial intelligence. We already know that simultaneously with the construction of artificial intelligence, we will have to delve not only into the human mind, but also into the essence of our society and in this context we have to consider tools that become social entities, which instinctively gives us the feeling of dealing with something uncanny [20]. REFERENCES [] Abu Shawar B., Atwell E. (2007) Chatbots: Are they Really Useful?, LDV-Forum Journal for Computational Linguistics and Language Technology, 22 (), pp [2] Abu Shawar B., Atwell E. (2004) Evaluation of Chatbot Information System, in Proceedings of the Eighth Maghrebian Conference on Software Engineering and Artificial Intelligence. [3] Banach A., Dąbkowski J. (200), Duży ruch to nie wsystko. Żeby gość nas polecał, Hotelar, (574) November, pp [4] Bartneck C., Rosalia C., Menges R., Deckers I. (2005) Robot Abuse - A Limitation of the Media Euation, in Proceedings of the INTERACT 2005 workshop Abuse: The darker side of Human-Computer Interaction, Rome, Italy. [5] Brahnam S. (2006) Gendered bods and bot abuse, in Proceedings of the CHI 2006 workshop on Misuse and abuse of interactive technologies, Montreal, Quebec, Canada. 56

75 [6] Brahnam S. (2005) Strategies for handling customer abuse of ECAs, in Proceedings of the INTERACT 2005 workshop Abuse: The darker side of Human-Computer Interaction, Rome, Italy, pp [7] Bogdanovych A., Simoff S., Sierra C., Berger H. (2005) Implicit training of virtual shopping assistants in 3D electronic institutions, in Proceedings of the IADIS International Conference: e-commerce 2005, IADIS Press, Portugal, pp [8] Chai J., Budikowska M., Horvath V., Nicolov N., Kambhatla N., Zadrony W. (200) Natural Language Sales Assistant - A Web-Based Dialog System for Online Sales, in Proceedings of the 3th Innovative Applications of Artificial Intelligence Conference, IAAI 0, Seattle, WA, pp [9] De Angeli A. (2006) On Verbal Abuse Towards Chatterbots, in Proceedings of the CHI 2006 workshop on Misuse and Abuse of Interactive Technologies, Montreal, Quebec, Canada. [0] De Angeli A., Brahnam S. (2008) I hate you! Disinhibition with virtual partners, Interacting with Computers, 20(3), pp [] De Angeli A., Carpenter R. (2005) Stupid computer! Abuse and social identities, in Proceedings of the INTERACT 2005 workshop Abuse: The darker side of Human- Computer Interaction, Rome, Italy. [2] De Angeli A., Johnson G. I., Coventry L. (200) The unfriendly user: exploring social reactions to chatterbots, in Proceedings of the International Conference on Affective Human Factor Design, London, pp [3] Hughes L. (2006) The Elia Effect: Conversational Agents and Cognition, available at: (accessed 27 September 20). [4] Jessa S. (2004) Cy chatterboty nas roumieją?, Software 2.0 Extra, 0/2004, pp. 6-20, available at: [5] Jessa S., Jędruch W. (200) Pretwaranie wyrażeń jęyka naturalnego w wyrażenia logicne - system Denise, in Predsięwięcia i usługi informacyjne. Praca biorowa Katedry Architektury Systemów Komputerowych KASKBOOK, (Ontologie w opisie scenariusy usług), Gdańsk, pp [6] Kuligowska K. (200) Kosty i koryści implementacji wirtualnych asystentów w predsiębiorstwach ora ich nacenie dla rowoju gospodarki elektronicnej, roprawa doktorska, Wydiał Nauk Ekonomicnych Uniwersytetu Warsawskiego, Warsawa. [7] Kuligowska K., Lasek M. (20) Virtual assistants support customer relations and business processes, The 0th International Conference on Information Management, Gdańsk. [8] Kopp S., Gesellensetter L., Krämer N., Wachsmuth I. (2004) A conversational agent as museum guide - design and evaluation of a real-world application, in Proceedings of Intelligent Virtual Agents (IVA 2005), Berlin, Germany, Volume 366, pp

76 [9] Loebner Prie, (20), available at: (accessed 27 September 20). [20] Mewes D., Heloir A. (2009) The Uncanny Valley, available at: (accessed 27 September 20). [2] Nass C., Steuer J., Tauber E. (994) Computers are social actors. Human Factors in Computing Systems, in CHI 94 Conference Proceedings, New York, pp [22] Pfeiffer T., Liguda C., Wachsmuth I., Stein S. (20) Living with a Virtual Agent: Seven Years with an Embodied Conversational Agent at the Hein Nixdorf MuseumsForum, in: S. Barbieri, K. Scott, & L. Ciolfi (eds.), Proceedings of the Re- Thinking Technology in Museums 20 - Emerging Experiences. Limmerick: think creative & the University of Limerick, pp [23] Reeves B. (2004) The Benefits of Interactive Online Characters, available at: (accessed 27 September 20). [24] Robinson S., Traum D., Ittycheriah M., Henderer J. (2008) What would you ask a conversational agent? observations of human-agent dialogues in a museum setting, in Proceedings of the 5th International Conference on Language Resources and Evaluation. [25] Saarine L. (200) Chatterbots: Crash Test Dummies of Communication. Master's Thesis, UIAH Helsinki, available at: (accessed 27 September 20). [26] Shieber S. (993) Lessons from a Restricted Turing Test, available at: (accessed 27 September 20). [27] Wallis, P. (2005) Believable conversational agents: introducing the intention map, available at: (accessed 27 Sep. 20). [28] Wallis P. (2005 b) Robust normative systems: what happens when a normative system fails?, in Proceedings of the INTERACT 2005 workshop Abuse: The darker side of Human-Computer Interaction, Rome, Italy. [29] Weienbaum J. (976) Computer power and human reason: from judgment to calculation, W. H. Freeman & Co., NY, USA. [30] Weienbaum J. (966) ELIZA A computer program for the study of natural language communication between man and machine, Communications of the ACM, 0(8), pp

77 INFORMATION SYSTEMS IN MANAGEMENT Information Systems in Management (203) Vol. 2 (2) 59 7 THE LEVEL OF LOGISTICS AND THE SOLUTIONS IN THE FIELD OF INFORMATION TECHNOLOGY APPLIED IN AGRIBUSINESS ENTERPRISES TOMASZ ROKICKI Department of Economics and Organisation of Enterprises, Warsaw University of Life Sciences (SGGW) The paper presents the relationship between the level of logistics and the solutions applied in the agribusiness enterprises. The research was conducted in 504 companies. The logistics level was determined on the basis of contribution of cost of logistics in the overall expenses of the company and based on existence of a separate logistics department in a given company. Relevant dependence has been detected only between existence of a separate organisational unit dedicated to logistics and the solutions applied in the area of logistics. For level of logistics costs, in most cases, no significant correlation with the use of information technology was seen. The share of cost of logistics in company expenses was determined to be a poor indicator of level of logistics and modernity of the company. Keywords: information systems, logistic, agribusiness. Introduction The logistics may be defined in many ways. It is a concept covering organisation, control and execution of flow of goods from place of manufacture to place of final purchase at minimum cost and minimum employment of funds [7]. According to Rutkowski the logistics deals with the management of operations of moving and storing in order to facilitate the flow of products and information from place of manufacture to the consumer at a reasonable cost [2]. According to Boarth and

78 Handfield the Logistics is a part of the supply chain, which deals with the planning, implementation and control of smooth and efficient flow of goods, services and information from the point of origin to point of consumption in order to meet the needs of the consumers [3]. A characteristic of all the definitions is not only smooth flow, but also avoiding the loss of time, uantity and funds. Logistics therefore relates to cover time and space in execution of flow of goods [6]. It enters into a number of relations with many functional areas, such as manufacturing, marketing and accounting [4]. The logistics, as a science, is in a continuous evolution, new theories are proposed and verified [3]. Of course, there are some classical assumptions, which remain unchanged. The level of logistics in a company may be measured in many ways. In this paper the author used cost of logistic activity and existence of one separate department dedicated to logistics to determine the level of logistics. Small enterprises are usually characterised by low level of logistics development [5]. The conducted research confirms relation between company sie and existence of a separate logistics department in its organisational structure. In small enterprises the presence of the logistics department is often unreasonable, whereas a large enterprise cannot operate effectively without such a department []. The logistic cost is more and more significant in the expenses borne by the enterprises. Its value depends mainly on the level of customer service and market competition. There are strong substitutive relationships between each type cost [0]. In 2008, as compared to 2003, logistics cost increased by 20%. A reason for this situation was increased cost connected with increased needs of the consumers. Reduction of expenses in the supply chains did not help too much [8]. In addition to smooth flow, the cost of the employed solution is a very important factor in assessment of the company logistics. The aim is to find economic balance between the expenses and robust handling of the material stream and service uality [2]. The information accompanies the flow of goods; it is the most important element building the efficient supply chain, and thus the customer satisfaction. Therefore the logistics of information should effective both inside as well as outside of the company [6]. Fulfilment of the assumed goals is possible using the information system, that is, a tool integrating the data coming from multiple sources into a single stream of information []. The software for optimising the production and the logistic processes, for conducting simulations as well as real time planning and scheduling, is offered on the market [4]. The agribusiness enterprises are still unexplored in terms of logistic solutions, including the solutions related to use of the information systems. The entities active in this sector differ from other production and service companies. In addition, the differences exist in each sector, for example, in terms of organisation of supplies, 60

79 model of relationships with the suppliers, seasonality of production, sales organisation, impact [9]. There are only a few detailed studies on management of logistics in agribusiness enterprises [5]. Therefore the subject taken is very important. 2. Methods of research The aim of the paper to determine dependence between the relationship between the level of logistics and the solutions applied in the enterprises active in the main sectors of the agribusiness. The data has been collected based on survey, made from December 2009 to March 200. The polls have been sent to all the enterprises classified as small, medium and large enterprises, operating in the food processing sector, existing in the REGON database and to randomly selected 500 micro-enterprises. The overall number was enterprises. 5 answers were received (5.%). The Dane for this study was used without processing, the analysis did not include only records with missing data, and thus 504 polls were considered. The structure of the companies by number of employees is shown in Figure. 6% 2% 20% 62% micro small medium large Figure. Structure of the enterprises surveyed The companies from the following sectors of the agribusiness were selected to the study: meat production and processing, fruit and vegetable processing, dairy production, production of grain mill and starch products, production of bakery and flour products, production of feed for animals, production of drinks. Primarily graphical analysis was used for data analysis. Due to properties of the data collect- 6

80 ed in nominal and ordinal scales. In order to determine the correlations between the selected traits also the chi suare statistical test was employed. 3. Research results The logistics level was determined on the basis of contribution of cost of logistics in the overall expenses of the company and based on presence of a separate department (person) dedicated to logistics. The enterprises unambiguously declared that they had or did not have a logistics department. Only 20% of the companies had such a department. This result demonstrates a general low level of logistics. Companies more often decided to separate organisational unit dealing with a selected area of logistics (fig. 2). Most often the companies established departments to deal with transport as well as storage. Such a result is not surprising, since for the companies the most important are production and distribution. Information management and packaging management units were the least rare ones in the agribusiness companies. 35,00% 30,00% 30,77% 29,06% 25,00% 20,00% 9,02% % 5,00% 0,00% 9,83%,32% 5,00% 0,00% transport supply management storage packaging management information management departments Figure 2. Functioning of separate logistics departments in agribusiness enterprises, multiple choice (%) The share of cost of logistics in the overall company expenses demonstrates the importance of logistics activities (fig. 3). Most often companies bore low logistics expenses (35% companies) or very low (30% companies). There were few units in which the expenses were very high. These results indicate low importance of logistics in the analysed enterprises. 62

81 % 40,0% 35,0% 30,0% 25,0% 20,0% 5,0% 0,0% 5,0% 0,0% 8,9% 25,8% 34,92% 30,37% very high high low very low share in overall company costs Figure 3. The share of logistics costs in the overall cost in agribusiness enterprises (%) Having a single comprehensive IT system allows the company to plan, coordinate and supervise its operations. Only 2% of the surveyed companies had a comprehensive system. Additional advantage to a comprehensive system is easy data transfer between different units of the enterprise. The surveyed enterprises made use of various IT systems. The most commonly used were financial and accounting systems FK (fig. 4). Such systems were used by 65% companies. MRP System (management of material and production resources) was used by 7% companies. ERP systems (enterprise resource management) were used less often than MRP (only 8% companies). Electronic data interchange (EDI) was used in 2% enterprises. Expert Decision Support System (BI Business Intelligence) was used by a few companies. Enterprises also made use of other systems (7% companies), but they did not disclose their names. The purchase of the specialised software is connected with high expenses, therefore the companies commonly use available finance & accounting systems. The performed chi 2 test for independence of variables showed dependence between the existence of a separate logistics department in the company and having a single, comprehensive IT system (chi 2 emp = 9,78, chi 2 0,05 = 3,84, p-value = 0.00, df = ). Whereas no dependence between cost of logistics and existence of a single IT system in the company was shown. The performed chi 2 test for independence of variables showed dependence between the existence of a separate logistics department in the company and using a MRP system (chi 2 emp = 0.78, chi 2 0,05 = 3.84, p-value < 0.05, df = ), ERP system (chi 2 emp =.93, p-value < 0,05), EDI (chi 2 emp = 30.39, p-value < 0.05). Nonetheless the hypothesis on lack of dependence between the existence of a dedicated logistics department and the use of accounting system was confirmed (chi 2 emp = 3.57, chi = 3.84, p-value < 0,05, df = ). A similar situation took 63

82 place in the case of the BI system (chi 2 emp = 2.2, p-value < 0.05). However no dependence between cost of logistics and use of each above-mentioned system was found (FK, MRP, ERP, EDI, BI). 70,00% 60,00% 65,7% % 50,00% 40,00% 30,00% 20,00% 0,00% 0,00% financial and accounting systems 7,22% 7,83%,55% 0,59% 6,5% MRP ERP EDI BI other IT systems Figure 4. The IT systems supporting the logistics of the agribusiness companies multiple choice (%) In the agribusiness sector enterprises usually storage management as well as order processing and stock management was supported by the IT systems (fig. 5). To a lesser extent modern IT solutions were used in the transport. Package management was the least supported. The results of the survey show that the companies employ IT solutions in key departments and in units, which are characterised by a high degree of complexity. The performed chi 2 tests for independence of variables showed lack of dependence between the existence of a separate logistics department and the IT supported order processing and forecasting (chi 2 emp = 3,82, p-value < 0.05) as well as the presence of strong dependence the IT supported transport (chi 2 emp = 24.09, p-value < 0.05), stock (chi 2 emp = 27.80, p-value < 0,05), packaging department and reverse logistics (chi 2 emp = 9.44, p-value < 0.05), stock management (chi 2 emp = 20,29, p-value < 0,05). The hypothesis of no relation between the level of logistics cost and the IT supported transport department was rejected (chi 2 emp = 5.87, chi 2 0,05 = 7.8, p-value = 0.05, df = 3). Similarly, in case of IT supported packaging and reverse logistics departments (chi 2 emp = 3.4, p-value < 0.05) and operations related to order processing and forecasting (chi 2 emp = 5.24, p-value < 0.05). Whereas small dependence between the level of logistics cost and aided stock management (chi 2 emp = 8.72, p-value < 0.05) and storage management (chi 2 emp = 0.5, p-value < 0.05). 64

83 % 45,00% 40,00% 35,00% 30,00% 25,00% 20,00% 5,00% 0,00% 5,00% 0,00% 27,0% 33,07% 4,29% transport supply packaging and reverse logistics areas of logistic opperations 36,59% stock management 38,94% order processing and forecasting Figure 5. The areas of IT supported logistic operations in the agribusiness companies - multiple choice (%) When dealing with external partners (suppliers and customers) information was transmitted mostly by telephone (fig. 6). 50% enterprises used or gave the instructions orally or in writing. Use of the records stored in the computer software was not popular. Only if the units are strongly related, the flow of information in such a way is possible. In external contacts traditional methods of communication were used, and the share of modern tools in the form of computer software was low. % 90,00% 80,00% 70,00% 60,00% 50,00% 40,00% 30,00% 20,00% 0,00% 0,00% 53,03% 46,58%,74% orally w riting computer softw are 85,52% 57,73%,57% telephone fax other w ay the way Figure 6. The way in which information in contact with external partners was delivered in agribusiness enterprises - multiple choice (%) The information inside the company was usually provided orally (fig. 7). The information was also provided by phone and text written down on the paper. More 65

84 modern methods of communication, such as , fax or storing in computer software, were used by very small percentage of the companies. Small businesses, because of the number of employees, used direct communication. In large companies more formal solutions were used. A small percentage of companies used computer for internal communication. 80,00% 75,5% % 70,00% 60,00% 50,00% 40,00% 38,6% 44,23% 30,00% 20,00% 0,00% 0,00% 22,90% 4,68% orally w riting computer softw are 4,68% 0,20% telephone fax other w ay the way Figure 7. The way in which information was delivered inside the company in agribusiness enterprises - multiple choice (%) % 50,00% 45,00% 40,00% 35,00% 30,00% 25,00% 20,00% 5,00% 0,00% 5,00% 0,00% 9,62% 44,37% 37,79% 8,22% leader over average below average one of w eakest assessment Figure 8. Assessment of the information management in agribusiness enterprises as compared to the competitors - multiple choice (%) IT support in the company was regarded as similar to other companies active in this sector. A small percentage of companies saw IT management in their enterprises as worse or even one of the worst in the sector. 66

85 4. Conclusion The agribusiness enterprises still focus on production activity, and to a lesser extent they see they opportunities in the implementation of the logistic solutions, including IT solutions. The low share of logistics cost in total cost jest confirms this regularity. The level of logistics in a company can be assessed either on the basis of the presence or absence of a separate department dedicated to logistics. The conducted research showed relation between presence of a separate logistics department and having a single comprehensive IT system or use of specialist systems, such as MRP, ERP, EDI. The use of modern information technology solutions is connected with huge initial expenses. In respect of the above, it could be afforded by large companies and, as the research showed, they were the companies aware of the importance of efficient logistics in the enterprise. Smaller companies commonly used finance & accounting systems, which serve for accounting operations and are necessary for the operation of the company. Whereas the BI systems were used very rarely. No actual relation between the level of logistics cost (measured by share in the total cost) and using modern IT solutions. The level of logistics costs does not show the scale of the enterprise activity. Generally small companies incur relatively high costs of logistics, and large companies purposefully incur expenses related to activity of the various areas of logistics. The information in companies either internal, or in contact with external partners is delivered via traditional channels. A few companies used computer software for communication. Logistics in agribusiness companies was at a low level, a similar situation was in case use of modern logistics solutions. The companies may benefit much from implementing logistic solutions in their activity, among others, by using modern IT tools. A limitation, especially for small businesses is the cost of purchasing and implementing these solutions. REFERENCES [] Baran J., Wysokiński M., Jałowiecki P. (20) Rowiąania w akresie sterowania apasami w wybranych branżach agrobinesu, Logistyka 2/20, Poland, [2] Beier F. J., Rutkowski K. (2000) Logistyka, Oficyna Wydawnica SGW, Warsawa, Poland. [3] Boharth C., Handfield R., B. (2007) Wprowadenie do arądania operacjami i łańcuchem dostaw, Wydawnictwo Helion, Gliwice, Poland. [4] Coyle J. J., Bardi E. J., Langley Jr C. J. (2007) Zarądanie logistycne, PWE, Warsawa, Poland,

86 [5] Klepacki B. (2008) Rowój logistyki jako cynnik wrostu predsiębiorstw agrobinesu, Rocniki Naukowe SERiA, t. X,. 3, Warsawa, Poland. [6] Logistyka (2009) red. nauk. Kisperska-Moroń D., Kryżaniak S., Wydawnictwo ILiM, Ponań, Poland, [7] Logistyka. Wybrane agadnienia (2008) Wydawnictwo SGGW, Warsawa, Poland, 9. [8] Pfohl H. Ch. (200) Doskonałość łańcucha dostaw w casach światowego kryysu gospodarcego, [ed.]: Logistyka wobec nowych wywań, Materiały konferencyjne, Biblioteka Logistyka, Ponań, Poland, [9] Rokicki T. (202) Organiacja logistyki w predsiębiorstwach premysłu mięsnego, Logistyka 2/202, CD nr 2, ILiM, Ponań, Poland, [0] Rokicki T., Wicki L. (200) Kosty logistyki w wybranych branżach agrobinesu, Prace Naukowe Uniwersytetu Ekonomicnego we Wrocławiu, Ekonomia, nr 3, Wydawnictwo Uniwersytetu Ekonomicnego we Wrocławiu. Wrocław, Poland, [] Rokicki T., Wicki L. (20) Supporting of main logistic processes by IT systems in enterprises of agri-food processing, Information Systems in management, e-logistics and Computer Aided Logistics. nr XV, Warsaw, WULS Press, Poland [2] Rokicki T., Wicki L. (20) The scope of the use of solutions to support logistics processes in the cereals processing companies, Information systems in management X, Computer aided logistics / sci. ed. Piotr Jałowiecki, Arkadius Orłowski, Warsaw, WULS Press, Poland, [3] Rynek usług logistycnych (2005) red. nauk. Ciesielski M., Difin, Warsawa, Poland, 0-. [4] Ważna L., Witkowski K. (2007) Komputerowe wspomaganie controllingu logistycnego wykorystaniem integrowanego systemu arądania proalpha APS w średnim predsiębiorstwie produkcyjnym [w:] Systemy informatycne w arądaniu, Wydawnictwo SGGW, Warsawa, Poland, 2-3. [5] Wicki L., Rokicki T. (20): Differentiation of level of logistics activities in milk processing companies, Information systems in management X: computer aided logistics / sci. ed. Piotr Jałowiecki, Arkadius Orłowski, Warsaw, WULS Press, Poland, [6] Wiktorowic J., Rosko-Gregorek E. (2008) Efektywne arądanie informacją w logistyce rola informatyki [w.]: Logistyka sansą rowoju miasta i regionu na prykładie iemi piotrkowskiej, Piotrków Trybunalski, Poland,